Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,8 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+checkpoint-1000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-1500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-1795/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,59 @@
+---
+library_name: peft
+license: other
+base_model: google/gemma-3-12b-it
+tags:
+- llama-factory
+- lora
+- generated_from_trainer
+model-index:
+- name: gemma3-12b-it_eqa_lora_sft
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# gemma3-12b-it_eqa_lora_sft
+
+This model is a fine-tuned version of [google/gemma-3-12b-it](https://huggingface.co/google/gemma-3-12b-it) on the energy_qa_alpaca dataset.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 0.0001
+- train_batch_size: 1
+- eval_batch_size: 8
+- seed: 42
+- gradient_accumulation_steps: 8
+- total_train_batch_size: 8
+- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 1.0
+
+### Training results
+
+
+
+### Framework versions
+
+- PEFT 0.15.2
+- Transformers 4.52.4
+- Pytorch 2.7.0+cu126
+- Datasets 3.6.0
+- Tokenizers 0.21.1

adapter_config.json

@@ -0,0 +1,180 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-3-12b-it",
+  "bias": "none",
+  "corda_config": null,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "language_model.layers.3.self_attn.q_proj",
+    "35.self_attn.k_proj",
+    "o_proj",
+    "40.self_attn.q_proj",
+    "language_model.layers.8.self_attn.k_proj",
+    "language_model.layers.25.self_attn.v_proj",
+    "language_model.layers.11.self_attn.k_proj",
+    "43.self_attn.q_proj",
+    "language_model.layers.15.self_attn.q_proj",
+    "39.self_attn.q_proj",
+    "language_model.layers.21.self_attn.v_proj",
+    "35.self_attn.q_proj",
+    "31.self_attn.v_proj",
+    "language_model.layers.8.self_attn.v_proj",
+    "38.self_attn.k_proj",
+    "language_model.layers.0.self_attn.k_proj",
+    "language_model.layers.12.self_attn.v_proj",
+    "27.self_attn.q_proj",
+    "language_model.layers.5.self_attn.q_proj",
+    "language_model.layers.2.self_attn.q_proj",
+    "language_model.layers.20.self_attn.k_proj",
+    "up_proj",
+    "language_model.layers.14.self_attn.q_proj",
+    "language_model.layers.3.self_attn.v_proj",
+    "language_model.layers.21.self_attn.q_proj",
+    "language_model.layers.26.self_attn.q_proj",
+    "down_proj",
+    "language_model.layers.23.self_attn.q_proj",
+    "language_model.layers.8.self_attn.q_proj",
+    "30.self_attn.v_proj",
+    "language_model.layers.1.self_attn.v_proj",
+    "36.self_attn.q_proj",
+    "28.self_attn.v_proj",
+    "language_model.layers.23.self_attn.v_proj",
+    "language_model.layers.22.self_attn.q_proj",
+    "language_model.layers.1.self_attn.k_proj",
+    "38.self_attn.q_proj",
+    "32.self_attn.q_proj",
+    "46.self_attn.k_proj",
+    "34.self_attn.k_proj",
+    "language_model.layers.16.self_attn.v_proj",
+    "language_model.layers.5.self_attn.v_proj",
+    "47.self_attn.k_proj",
+    "42.self_attn.k_proj",
+    "45.self_attn.k_proj",
+    "45.self_attn.v_proj",
+    "language_model.layers.19.self_attn.k_proj",
+    "language_model.layers.22.self_attn.k_proj",
+    "language_model.layers.17.self_attn.v_proj",
+    "language_model.layers.17.self_attn.q_proj",
+    "language_model.layers.9.self_attn.k_proj",
+    "language_model.layers.6.self_attn.q_proj",
+    "language_model.layers.1.self_attn.q_proj",
+    "language_model.layers.7.self_attn.k_proj",
+    "39.self_attn.k_proj",
+    "language_model.layers.6.self_attn.k_proj",
+    "language_model.layers.18.self_attn.q_proj",
+    "language_model.layers.5.self_attn.k_proj",
+    "29.self_attn.k_proj",
+    "language_model.layers.20.self_attn.v_proj",
+    "language_model.layers.25.self_attn.k_proj",
+    "language_model.layers.4.self_attn.q_proj",
+    "language_model.layers.13.self_attn.k_proj",
+    "41.self_attn.v_proj",
+    "language_model.layers.12.self_attn.q_proj",
+    "language_model.layers.7.self_attn.q_proj",
+    "language_model.layers.19.self_attn.v_proj",
+    "language_model.layers.7.self_attn.v_proj",
+    "language_model.layers.11.self_attn.q_proj",
+    "30.self_attn.k_proj",
+    "language_model.layers.16.self_attn.k_proj",
+    "43.self_attn.k_proj",
+    "47.self_attn.q_proj",
+    "language_model.layers.10.self_attn.v_proj",
+    "language_model.layers.15.self_attn.v_proj",
+    "language_model.layers.0.self_attn.v_proj",
+    "31.self_attn.k_proj",
+    "35.self_attn.v_proj",
+    "language_model.layers.2.self_attn.k_proj",
+    "27.self_attn.v_proj",
+    "30.self_attn.q_proj",
+    "28.self_attn.k_proj",
+    "41.self_attn.k_proj",
+    "language_model.layers.0.self_attn.q_proj",
+    "33.self_attn.v_proj",
+    "language_model.layers.6.self_attn.v_proj",
+    "44.self_attn.q_proj",
+    "language_model.layers.2.self_attn.v_proj",
+    "language_model.layers.3.self_attn.k_proj",
+    "language_model.layers.9.self_attn.q_proj",
+    "37.self_attn.q_proj",
+    "language_model.layers.4.self_attn.v_proj",
+    "33.self_attn.q_proj",
+    "29.self_attn.v_proj",
+    "language_model.layers.18.self_attn.k_proj",
+    "37.self_attn.v_proj",
+    "language_model.layers.17.self_attn.k_proj",
+    "33.self_attn.k_proj",
+    "language_model.layers.19.self_attn.q_proj",
+    "gate_proj",
+    "language_model.layers.10.self_attn.q_proj",
+    "47.self_attn.v_proj",
+    "42.self_attn.q_proj",
+    "language_model.layers.16.self_attn.q_proj",
+    "language_model.layers.25.self_attn.q_proj",
+    "42.self_attn.v_proj",
+    "32.self_attn.k_proj",
+    "language_model.layers.9.self_attn.v_proj",
+    "46.self_attn.v_proj",
+    "language_model.layers.12.self_attn.k_proj",
+    "38.self_attn.v_proj",
+    "language_model.layers.14.self_attn.k_proj",
+    "46.self_attn.q_proj",
+    "language_model.layers.18.self_attn.v_proj",
+    "language_model.layers.26.self_attn.v_proj",
+    "44.self_attn.v_proj",
+    "language_model.layers.24.self_attn.q_proj",
+    "language_model.layers.15.self_attn.k_proj",
+    "language_model.layers.14.self_attn.v_proj",
+    "29.self_attn.q_proj",
+    "language_model.layers.4.self_attn.k_proj",
+    "language_model.layers.13.self_attn.q_proj",
+    "43.self_attn.v_proj",
+    "34.self_attn.q_proj",
+    "45.self_attn.q_proj",
+    "32.self_attn.v_proj",
+    "language_model.layers.11.self_attn.v_proj",
+    "36.self_attn.v_proj",
+    "language_model.layers.21.self_attn.k_proj",
+    "31.self_attn.q_proj",
+    "37.self_attn.k_proj",
+    "39.self_attn.v_proj",
+    "language_model.layers.26.self_attn.k_proj",
+    "40.self_attn.k_proj",
+    "36.self_attn.k_proj",
+    "41.self_attn.q_proj",
+    "language_model.layers.24.self_attn.v_proj",
+    "language_model.layers.23.self_attn.k_proj",
+    "language_model.layers.10.self_attn.k_proj",
+    "27.self_attn.k_proj",
+    "language_model.layers.13.self_attn.v_proj",
+    "34.self_attn.v_proj",
+    "40.self_attn.v_proj",
+    "language_model.layers.20.self_attn.q_proj",
+    "language_model.layers.24.self_attn.k_proj",
+    "44.self_attn.k_proj",
+    "28.self_attn.q_proj",
+    "language_model.layers.22.self_attn.v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f605bf90a7e13b22d18038eeb0bffe63d3b68562cec8e726a11e3f9fa51433d
+size 131040264

added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<image_soft_token>": 262144
+}

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 1.0,
+    "total_flos": 4.2177966191161344e+17,
+    "train_loss": 1.5585720649336705,
+    "train_runtime": 12848.5461,
+    "train_samples_per_second": 1.117,
+    "train_steps_per_second": 0.14
+}

chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

checkpoint-1000/README.md

@@ -0,0 +1,202 @@
+---
+base_model: google/gemma-3-12b-it
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.15.2

checkpoint-1000/adapter_config.json

@@ -0,0 +1,180 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-3-12b-it",
+  "bias": "none",
+  "corda_config": null,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "language_model.layers.3.self_attn.q_proj",
+    "35.self_attn.k_proj",
+    "o_proj",
+    "40.self_attn.q_proj",
+    "language_model.layers.8.self_attn.k_proj",
+    "language_model.layers.25.self_attn.v_proj",
+    "language_model.layers.11.self_attn.k_proj",
+    "43.self_attn.q_proj",
+    "language_model.layers.15.self_attn.q_proj",
+    "39.self_attn.q_proj",
+    "language_model.layers.21.self_attn.v_proj",
+    "35.self_attn.q_proj",
+    "31.self_attn.v_proj",
+    "language_model.layers.8.self_attn.v_proj",
+    "38.self_attn.k_proj",
+    "language_model.layers.0.self_attn.k_proj",
+    "language_model.layers.12.self_attn.v_proj",
+    "27.self_attn.q_proj",
+    "language_model.layers.5.self_attn.q_proj",
+    "language_model.layers.2.self_attn.q_proj",
+    "language_model.layers.20.self_attn.k_proj",
+    "up_proj",
+    "language_model.layers.14.self_attn.q_proj",
+    "language_model.layers.3.self_attn.v_proj",
+    "language_model.layers.21.self_attn.q_proj",
+    "language_model.layers.26.self_attn.q_proj",
+    "down_proj",
+    "language_model.layers.23.self_attn.q_proj",
+    "language_model.layers.8.self_attn.q_proj",
+    "30.self_attn.v_proj",
+    "language_model.layers.1.self_attn.v_proj",
+    "36.self_attn.q_proj",
+    "28.self_attn.v_proj",
+    "language_model.layers.23.self_attn.v_proj",
+    "language_model.layers.22.self_attn.q_proj",
+    "language_model.layers.1.self_attn.k_proj",
+    "38.self_attn.q_proj",
+    "32.self_attn.q_proj",
+    "46.self_attn.k_proj",
+    "34.self_attn.k_proj",
+    "language_model.layers.16.self_attn.v_proj",
+    "language_model.layers.5.self_attn.v_proj",
+    "47.self_attn.k_proj",
+    "42.self_attn.k_proj",
+    "45.self_attn.k_proj",
+    "45.self_attn.v_proj",
+    "language_model.layers.19.self_attn.k_proj",
+    "language_model.layers.22.self_attn.k_proj",
+    "language_model.layers.17.self_attn.v_proj",
+    "language_model.layers.17.self_attn.q_proj",
+    "language_model.layers.9.self_attn.k_proj",
+    "language_model.layers.6.self_attn.q_proj",
+    "language_model.layers.1.self_attn.q_proj",
+    "language_model.layers.7.self_attn.k_proj",
+    "39.self_attn.k_proj",
+    "language_model.layers.6.self_attn.k_proj",
+    "language_model.layers.18.self_attn.q_proj",
+    "language_model.layers.5.self_attn.k_proj",
+    "29.self_attn.k_proj",
+    "language_model.layers.20.self_attn.v_proj",
+    "language_model.layers.25.self_attn.k_proj",
+    "language_model.layers.4.self_attn.q_proj",
+    "language_model.layers.13.self_attn.k_proj",
+    "41.self_attn.v_proj",
+    "language_model.layers.12.self_attn.q_proj",
+    "language_model.layers.7.self_attn.q_proj",
+    "language_model.layers.19.self_attn.v_proj",
+    "language_model.layers.7.self_attn.v_proj",
+    "language_model.layers.11.self_attn.q_proj",
+    "30.self_attn.k_proj",
+    "language_model.layers.16.self_attn.k_proj",
+    "43.self_attn.k_proj",
+    "47.self_attn.q_proj",
+    "language_model.layers.10.self_attn.v_proj",
+    "language_model.layers.15.self_attn.v_proj",
+    "language_model.layers.0.self_attn.v_proj",
+    "31.self_attn.k_proj",
+    "35.self_attn.v_proj",
+    "language_model.layers.2.self_attn.k_proj",
+    "27.self_attn.v_proj",
+    "30.self_attn.q_proj",
+    "28.self_attn.k_proj",
+    "41.self_attn.k_proj",
+    "language_model.layers.0.self_attn.q_proj",
+    "33.self_attn.v_proj",
+    "language_model.layers.6.self_attn.v_proj",
+    "44.self_attn.q_proj",
+    "language_model.layers.2.self_attn.v_proj",
+    "language_model.layers.3.self_attn.k_proj",
+    "language_model.layers.9.self_attn.q_proj",
+    "37.self_attn.q_proj",
+    "language_model.layers.4.self_attn.v_proj",
+    "33.self_attn.q_proj",
+    "29.self_attn.v_proj",
+    "language_model.layers.18.self_attn.k_proj",
+    "37.self_attn.v_proj",
+    "language_model.layers.17.self_attn.k_proj",
+    "33.self_attn.k_proj",
+    "language_model.layers.19.self_attn.q_proj",
+    "gate_proj",
+    "language_model.layers.10.self_attn.q_proj",
+    "47.self_attn.v_proj",
+    "42.self_attn.q_proj",
+    "language_model.layers.16.self_attn.q_proj",
+    "language_model.layers.25.self_attn.q_proj",
+    "42.self_attn.v_proj",
+    "32.self_attn.k_proj",
+    "language_model.layers.9.self_attn.v_proj",
+    "46.self_attn.v_proj",
+    "language_model.layers.12.self_attn.k_proj",
+    "38.self_attn.v_proj",
+    "language_model.layers.14.self_attn.k_proj",
+    "46.self_attn.q_proj",
+    "language_model.layers.18.self_attn.v_proj",
+    "language_model.layers.26.self_attn.v_proj",
+    "44.self_attn.v_proj",
+    "language_model.layers.24.self_attn.q_proj",
+    "language_model.layers.15.self_attn.k_proj",
+    "language_model.layers.14.self_attn.v_proj",
+    "29.self_attn.q_proj",
+    "language_model.layers.4.self_attn.k_proj",
+    "language_model.layers.13.self_attn.q_proj",
+    "43.self_attn.v_proj",
+    "34.self_attn.q_proj",
+    "45.self_attn.q_proj",
+    "32.self_attn.v_proj",
+    "language_model.layers.11.self_attn.v_proj",
+    "36.self_attn.v_proj",
+    "language_model.layers.21.self_attn.k_proj",
+    "31.self_attn.q_proj",
+    "37.self_attn.k_proj",
+    "39.self_attn.v_proj",
+    "language_model.layers.26.self_attn.k_proj",
+    "40.self_attn.k_proj",
+    "36.self_attn.k_proj",
+    "41.self_attn.q_proj",
+    "language_model.layers.24.self_attn.v_proj",
+    "language_model.layers.23.self_attn.k_proj",
+    "language_model.layers.10.self_attn.k_proj",
+    "27.self_attn.k_proj",
+    "language_model.layers.13.self_attn.v_proj",
+    "34.self_attn.v_proj",
+    "40.self_attn.v_proj",
+    "language_model.layers.20.self_attn.q_proj",
+    "language_model.layers.24.self_attn.k_proj",
+    "44.self_attn.k_proj",
+    "28.self_attn.q_proj",
+    "language_model.layers.22.self_attn.v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-1000/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcc4a63960425bdb5a90c94400d2a37e71710f399f245faa2f15d8528ff9ff59
+size 131040264

checkpoint-1000/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<image_soft_token>": 262144
+}

checkpoint-1000/chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

checkpoint-1000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11f6cddd7633265ba463db0f5eda4b491f26487597f6dcc2be68a2799138b6df
+size 262448707

checkpoint-1000/preprocessor_config.json

@@ -0,0 +1,29 @@
+{
+  "do_convert_rgb": null,
+  "do_normalize": true,
+  "do_pan_and_scan": null,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "Gemma3ImageProcessor",
+  "image_seq_length": 256,
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "pan_and_scan_max_num_crops": null,
+  "pan_and_scan_min_crop_size": null,
+  "pan_and_scan_min_ratio_to_activate": null,
+  "processor_class": "Gemma3Processor",
+  "resample": 2,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "height": 896,
+    "width": 896
+  }
+}

checkpoint-1000/processor_config.json

@@ -0,0 +1,4 @@
+{
+  "image_seq_length": 256,
+  "processor_class": "Gemma3Processor"
+}

checkpoint-1000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:250560ab3d528161ab3659b120def6e4a9ab4b457e3399603bbcfa40db3efc90
+size 14645

checkpoint-1000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7742f57d6748a994796e54556d7aa815c4953a13a94d3bf47d4e1c767c576e9
+size 1465

checkpoint-1000/special_tokens_map.json

@@ -0,0 +1,33 @@
+{
+  "boi_token": "<start_of_image>",
+  "bos_token": {
+    "content": "<bos>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eoi_token": "<end_of_image>",
+  "eos_token": {
+    "content": "<end_of_turn>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "image_token": "<image_soft_token>",
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-1000/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4667f2089529e8e7657cfb6d1c19910ae71ff5f28aa7ab2ff2763330affad795
+size 33384568

checkpoint-1000/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c
+size 4689074

checkpoint-1000/trainer_state.json

@@ -0,0 +1,1434 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.557374764857521,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0027868738242876052,
+      "grad_norm": 2.1914708614349365,
+      "learning_rate": 2.2222222222222225e-06,
+      "loss": 2.8194,
+      "step": 5
+    },
+    {
+      "epoch": 0.0055737476485752105,
+      "grad_norm": 2.056274890899658,
+      "learning_rate": 5e-06,
+      "loss": 2.7461,
+      "step": 10
+    },
+    {
+      "epoch": 0.008360621472862817,
+      "grad_norm": 1.9419809579849243,
+      "learning_rate": 7.777777777777777e-06,
+      "loss": 2.7894,
+      "step": 15
+    },
+    {
+      "epoch": 0.011147495297150421,
+      "grad_norm": 2.0836448669433594,
+      "learning_rate": 1.0555555555555555e-05,
+      "loss": 2.8633,
+      "step": 20
+    },
+    {
+      "epoch": 0.013934369121438027,
+      "grad_norm": 2.413074493408203,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 2.7385,
+      "step": 25
+    },
+    {
+      "epoch": 0.016721242945725634,
+      "grad_norm": 2.3652689456939697,
+      "learning_rate": 1.6111111111111115e-05,
+      "loss": 2.6383,
+      "step": 30
+    },
+    {
+      "epoch": 0.019508116770013236,
+      "grad_norm": 1.4635894298553467,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 2.3059,
+      "step": 35
+    },
+    {
+      "epoch": 0.022294990594300842,
+      "grad_norm": 1.319265365600586,
+      "learning_rate": 2.1666666666666667e-05,
+      "loss": 2.0529,
+      "step": 40
+    },
+    {
+      "epoch": 0.025081864418588447,
+      "grad_norm": 1.2513976097106934,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 2.0179,
+      "step": 45
+    },
+    {
+      "epoch": 0.027868738242876053,
+      "grad_norm": 1.264099359512329,
+      "learning_rate": 2.7222222222222223e-05,
+      "loss": 1.9416,
+      "step": 50
+    },
+    {
+      "epoch": 0.03065561206716366,
+      "grad_norm": 1.0929852724075317,
+      "learning_rate": 3e-05,
+      "loss": 1.7526,
+      "step": 55
+    },
+    {
+      "epoch": 0.03344248589145127,
+      "grad_norm": 1.0694290399551392,
+      "learning_rate": 3.277777777777778e-05,
+      "loss": 1.9019,
+      "step": 60
+    },
+    {
+      "epoch": 0.03622935971573887,
+      "grad_norm": 1.2505921125411987,
+      "learning_rate": 3.555555555555556e-05,
+      "loss": 1.7937,
+      "step": 65
+    },
+    {
+      "epoch": 0.03901623354002647,
+      "grad_norm": 1.1209263801574707,
+      "learning_rate": 3.8333333333333334e-05,
+      "loss": 1.6891,
+      "step": 70
+    },
+    {
+      "epoch": 0.04180310736431408,
+      "grad_norm": 1.1830799579620361,
+      "learning_rate": 4.111111111111111e-05,
+      "loss": 1.7494,
+      "step": 75
+    },
+    {
+      "epoch": 0.044589981188601684,
+      "grad_norm": 1.0991789102554321,
+      "learning_rate": 4.388888888888889e-05,
+      "loss": 1.6545,
+      "step": 80
+    },
+    {
+      "epoch": 0.04737685501288929,
+      "grad_norm": 1.075130820274353,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 1.7803,
+      "step": 85
+    },
+    {
+      "epoch": 0.050163728837176895,
+      "grad_norm": 1.1547001600265503,
+      "learning_rate": 4.9444444444444446e-05,
+      "loss": 1.6842,
+      "step": 90
+    },
+    {
+      "epoch": 0.0529506026614645,
+      "grad_norm": 1.2314906120300293,
+      "learning_rate": 5.222222222222223e-05,
+      "loss": 1.6707,
+      "step": 95
+    },
+    {
+      "epoch": 0.055737476485752106,
+      "grad_norm": 1.1287559270858765,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 1.6046,
+      "step": 100
+    },
+    {
+      "epoch": 0.05852435031003971,
+      "grad_norm": 1.0470261573791504,
+      "learning_rate": 5.7777777777777776e-05,
+      "loss": 1.6453,
+      "step": 105
+    },
+    {
+      "epoch": 0.06131122413432732,
+      "grad_norm": 1.3171180486679077,
+      "learning_rate": 6.055555555555555e-05,
+      "loss": 1.7238,
+      "step": 110
+    },
+    {
+      "epoch": 0.06409809795861493,
+      "grad_norm": 1.1685153245925903,
+      "learning_rate": 6.333333333333333e-05,
+      "loss": 1.6259,
+      "step": 115
+    },
+    {
+      "epoch": 0.06688497178290254,
+      "grad_norm": 1.006691813468933,
+      "learning_rate": 6.611111111111111e-05,
+      "loss": 1.6396,
+      "step": 120
+    },
+    {
+      "epoch": 0.06967184560719013,
+      "grad_norm": 1.1026668548583984,
+      "learning_rate": 6.88888888888889e-05,
+      "loss": 1.7253,
+      "step": 125
+    },
+    {
+      "epoch": 0.07245871943147773,
+      "grad_norm": 1.1105059385299683,
+      "learning_rate": 7.166666666666667e-05,
+      "loss": 1.6867,
+      "step": 130
+    },
+    {
+      "epoch": 0.07524559325576534,
+      "grad_norm": 1.0630534887313843,
+      "learning_rate": 7.444444444444444e-05,
+      "loss": 1.6371,
+      "step": 135
+    },
+    {
+      "epoch": 0.07803246708005294,
+      "grad_norm": 1.0416017770767212,
+      "learning_rate": 7.722222222222223e-05,
+      "loss": 1.643,
+      "step": 140
+    },
+    {
+      "epoch": 0.08081934090434055,
+      "grad_norm": 0.9604218006134033,
+      "learning_rate": 8e-05,
+      "loss": 1.7327,
+      "step": 145
+    },
+    {
+      "epoch": 0.08360621472862816,
+      "grad_norm": 1.033799648284912,
+      "learning_rate": 8.277777777777778e-05,
+      "loss": 1.6771,
+      "step": 150
+    },
+    {
+      "epoch": 0.08639308855291576,
+      "grad_norm": 0.9786093831062317,
+      "learning_rate": 8.555555555555556e-05,
+      "loss": 1.6525,
+      "step": 155
+    },
+    {
+      "epoch": 0.08917996237720337,
+      "grad_norm": 1.0249221324920654,
+      "learning_rate": 8.833333333333333e-05,
+      "loss": 1.6163,
+      "step": 160
+    },
+    {
+      "epoch": 0.09196683620149097,
+      "grad_norm": 1.0517358779907227,
+      "learning_rate": 9.111111111111112e-05,
+      "loss": 1.5926,
+      "step": 165
+    },
+    {
+      "epoch": 0.09475371002577858,
+      "grad_norm": 1.0581194162368774,
+      "learning_rate": 9.388888888888889e-05,
+      "loss": 1.6449,
+      "step": 170
+    },
+    {
+      "epoch": 0.09754058385006618,
+      "grad_norm": 1.0320991277694702,
+      "learning_rate": 9.666666666666667e-05,
+      "loss": 1.6547,
+      "step": 175
+    },
+    {
+      "epoch": 0.10032745767435379,
+      "grad_norm": 1.012792706489563,
+      "learning_rate": 9.944444444444446e-05,
+      "loss": 1.6617,
+      "step": 180
+    },
+    {
+      "epoch": 0.1031143314986414,
+      "grad_norm": 1.0023682117462158,
+      "learning_rate": 9.999848639521432e-05,
+      "loss": 1.598,
+      "step": 185
+    },
+    {
+      "epoch": 0.105901205322929,
+      "grad_norm": 1.0282474756240845,
+      "learning_rate": 9.999233753283091e-05,
+      "loss": 1.5899,
+      "step": 190
+    },
+    {
+      "epoch": 0.1086880791472166,
+      "grad_norm": 1.2141717672348022,
+      "learning_rate": 9.998145939378577e-05,
+      "loss": 1.585,
+      "step": 195
+    },
+    {
+      "epoch": 0.11147495297150421,
+      "grad_norm": 0.9987207651138306,
+      "learning_rate": 9.996585300715116e-05,
+      "loss": 1.6266,
+      "step": 200
+    },
+    {
+      "epoch": 0.11426182679579182,
+      "grad_norm": 0.9829638600349426,
+      "learning_rate": 9.994551984929175e-05,
+      "loss": 1.6098,
+      "step": 205
+    },
+    {
+      "epoch": 0.11704870062007942,
+      "grad_norm": 0.8391382098197937,
+      "learning_rate": 9.992046184372492e-05,
+      "loss": 1.628,
+      "step": 210
+    },
+    {
+      "epoch": 0.11983557444436703,
+      "grad_norm": 0.8903898596763611,
+      "learning_rate": 9.989068136093873e-05,
+      "loss": 1.5645,
+      "step": 215
+    },
+    {
+      "epoch": 0.12262244826865464,
+      "grad_norm": 0.9336121082305908,
+      "learning_rate": 9.985618121816779e-05,
+      "loss": 1.6621,
+      "step": 220
+    },
+    {
+      "epoch": 0.12540932209294225,
+      "grad_norm": 0.9730141162872314,
+      "learning_rate": 9.981696467912664e-05,
+      "loss": 1.5838,
+      "step": 225
+    },
+    {
+      "epoch": 0.12819619591722986,
+      "grad_norm": 0.9329681992530823,
+      "learning_rate": 9.97730354537011e-05,
+      "loss": 1.6099,
+      "step": 230
+    },
+    {
+      "epoch": 0.13098306974151747,
+      "grad_norm": 0.8763962984085083,
+      "learning_rate": 9.972439769759722e-05,
+      "loss": 1.516,
+      "step": 235
+    },
+    {
+      "epoch": 0.13376994356580507,
+      "grad_norm": 0.9208707213401794,
+      "learning_rate": 9.967105601194823e-05,
+      "loss": 1.6798,
+      "step": 240
+    },
+    {
+      "epoch": 0.13655681739009268,
+      "grad_norm": 0.9264675378799438,
+      "learning_rate": 9.961301544287922e-05,
+      "loss": 1.545,
+      "step": 245
+    },
+    {
+      "epoch": 0.13934369121438026,
+      "grad_norm": 1.0418734550476074,
+      "learning_rate": 9.955028148102979e-05,
+      "loss": 1.5595,
+      "step": 250
+    },
+    {
+      "epoch": 0.14213056503866786,
+      "grad_norm": 0.9690463542938232,
+      "learning_rate": 9.948286006103466e-05,
+      "loss": 1.6868,
+      "step": 255
+    },
+    {
+      "epoch": 0.14491743886295547,
+      "grad_norm": 0.9389734864234924,
+      "learning_rate": 9.941075756096226e-05,
+      "loss": 1.6089,
+      "step": 260
+    },
+    {
+      "epoch": 0.14770431268724307,
+      "grad_norm": 0.902563750743866,
+      "learning_rate": 9.933398080171123e-05,
+      "loss": 1.6093,
+      "step": 265
+    },
+    {
+      "epoch": 0.15049118651153068,
+      "grad_norm": 0.9173883199691772,
+      "learning_rate": 9.925253704636543e-05,
+      "loss": 1.5829,
+      "step": 270
+    },
+    {
+      "epoch": 0.15327806033581828,
+      "grad_norm": 0.9028013944625854,
+      "learning_rate": 9.916643399950656e-05,
+      "loss": 1.5999,
+      "step": 275
+    },
+    {
+      "epoch": 0.1560649341601059,
+      "grad_norm": 0.9312839508056641,
+      "learning_rate": 9.907567980648549e-05,
+      "loss": 1.5684,
+      "step": 280
+    },
+    {
+      "epoch": 0.1588518079843935,
+      "grad_norm": 0.8616472482681274,
+      "learning_rate": 9.898028305265169e-05,
+      "loss": 1.6562,
+      "step": 285
+    },
+    {
+      "epoch": 0.1616386818086811,
+      "grad_norm": 0.8607208728790283,
+      "learning_rate": 9.888025276254096e-05,
+      "loss": 1.6552,
+      "step": 290
+    },
+    {
+      "epoch": 0.1644255556329687,
+      "grad_norm": 0.8621200323104858,
+      "learning_rate": 9.877559839902184e-05,
+      "loss": 1.5795,
+      "step": 295
+    },
+    {
+      "epoch": 0.1672124294572563,
+      "grad_norm": 0.9334512948989868,
+      "learning_rate": 9.86663298624003e-05,
+      "loss": 1.5635,
+      "step": 300
+    },
+    {
+      "epoch": 0.16999930328154392,
+      "grad_norm": 0.9203838109970093,
+      "learning_rate": 9.855245748948326e-05,
+      "loss": 1.6554,
+      "step": 305
+    },
+    {
+      "epoch": 0.17278617710583152,
+      "grad_norm": 0.9085136651992798,
+      "learning_rate": 9.843399205260068e-05,
+      "loss": 1.6244,
+      "step": 310
+    },
+    {
+      "epoch": 0.17557305093011913,
+      "grad_norm": 1.005462884902954,
+      "learning_rate": 9.831094475858652e-05,
+      "loss": 1.6093,
+      "step": 315
+    },
+    {
+      "epoch": 0.17835992475440673,
+      "grad_norm": 0.9284034371376038,
+      "learning_rate": 9.818332724771857e-05,
+      "loss": 1.6564,
+      "step": 320
+    },
+    {
+      "epoch": 0.18114679857869434,
+      "grad_norm": 0.8924919962882996,
+      "learning_rate": 9.805115159261726e-05,
+      "loss": 1.5466,
+      "step": 325
+    },
+    {
+      "epoch": 0.18393367240298195,
+      "grad_norm": 0.8804652094841003,
+      "learning_rate": 9.791443029710361e-05,
+      "loss": 1.5292,
+      "step": 330
+    },
+    {
+      "epoch": 0.18672054622726955,
+      "grad_norm": 0.9378550052642822,
+      "learning_rate": 9.777317629501636e-05,
+      "loss": 1.5443,
+      "step": 335
+    },
+    {
+      "epoch": 0.18950742005155716,
+      "grad_norm": 0.812350869178772,
+      "learning_rate": 9.762740294898846e-05,
+      "loss": 1.5034,
+      "step": 340
+    },
+    {
+      "epoch": 0.19229429387584476,
+      "grad_norm": 0.871232271194458,
+      "learning_rate": 9.747712404918286e-05,
+      "loss": 1.4828,
+      "step": 345
+    },
+    {
+      "epoch": 0.19508116770013237,
+      "grad_norm": 1.0012617111206055,
+      "learning_rate": 9.732235381198813e-05,
+      "loss": 1.6139,
+      "step": 350
+    },
+    {
+      "epoch": 0.19786804152441997,
+      "grad_norm": 0.8306112289428711,
+      "learning_rate": 9.716310687867342e-05,
+      "loss": 1.4516,
+      "step": 355
+    },
+    {
+      "epoch": 0.20065491534870758,
+      "grad_norm": 0.925584614276886,
+      "learning_rate": 9.699939831400351e-05,
+      "loss": 1.6499,
+      "step": 360
+    },
+    {
+      "epoch": 0.20344178917299519,
+      "grad_norm": 0.9054883718490601,
+      "learning_rate": 9.683124360481364e-05,
+      "loss": 1.6033,
+      "step": 365
+    },
+    {
+      "epoch": 0.2062286629972828,
+      "grad_norm": 0.9771924614906311,
+      "learning_rate": 9.665865865854445e-05,
+      "loss": 1.5691,
+      "step": 370
+    },
+    {
+      "epoch": 0.2090155368215704,
+      "grad_norm": 0.9042426347732544,
+      "learning_rate": 9.648165980173712e-05,
+      "loss": 1.5722,
+      "step": 375
+    },
+    {
+      "epoch": 0.211802410645858,
+      "grad_norm": 0.9686840772628784,
+      "learning_rate": 9.630026377848892e-05,
+      "loss": 1.5936,
+      "step": 380
+    },
+    {
+      "epoch": 0.2145892844701456,
+      "grad_norm": 0.8492907285690308,
+      "learning_rate": 9.611448774886924e-05,
+      "loss": 1.5515,
+      "step": 385
+    },
+    {
+      "epoch": 0.2173761582944332,
+      "grad_norm": 0.95368891954422,
+      "learning_rate": 9.592434928729616e-05,
+      "loss": 1.6242,
+      "step": 390
+    },
+    {
+      "epoch": 0.22016303211872082,
+      "grad_norm": 0.9186941981315613,
+      "learning_rate": 9.572986638087396e-05,
+      "loss": 1.5976,
+      "step": 395
+    },
+    {
+      "epoch": 0.22294990594300843,
+      "grad_norm": 0.8386760354042053,
+      "learning_rate": 9.553105742769154e-05,
+      "loss": 1.5469,
+      "step": 400
+    },
+    {
+      "epoch": 0.22573677976729603,
+      "grad_norm": 0.875751793384552,
+      "learning_rate": 9.532794123508197e-05,
+      "loss": 1.4213,
+      "step": 405
+    },
+    {
+      "epoch": 0.22852365359158364,
+      "grad_norm": 0.8671753406524658,
+      "learning_rate": 9.512053701784329e-05,
+      "loss": 1.5776,
+      "step": 410
+    },
+    {
+      "epoch": 0.23131052741587124,
+      "grad_norm": 0.8599863052368164,
+      "learning_rate": 9.490886439642081e-05,
+      "loss": 1.5758,
+      "step": 415
+    },
+    {
+      "epoch": 0.23409740124015885,
+      "grad_norm": 0.9010843634605408,
+      "learning_rate": 9.469294339505098e-05,
+      "loss": 1.6157,
+      "step": 420
+    },
+    {
+      "epoch": 0.23688427506444645,
+      "grad_norm": 0.8823115229606628,
+      "learning_rate": 9.447279443986716e-05,
+      "loss": 1.5334,
+      "step": 425
+    },
+    {
+      "epoch": 0.23967114888873406,
+      "grad_norm": 0.9563323855400085,
+      "learning_rate": 9.424843835696724e-05,
+      "loss": 1.5605,
+      "step": 430
+    },
+    {
+      "epoch": 0.24245802271302166,
+      "grad_norm": 0.9097768068313599,
+      "learning_rate": 9.401989637044355e-05,
+      "loss": 1.5316,
+      "step": 435
+    },
+    {
+      "epoch": 0.24524489653730927,
+      "grad_norm": 0.8611193299293518,
+      "learning_rate": 9.3787190100375e-05,
+      "loss": 1.5611,
+      "step": 440
+    },
+    {
+      "epoch": 0.24803177036159688,
+      "grad_norm": 0.8489307761192322,
+      "learning_rate": 9.355034156078188e-05,
+      "loss": 1.5352,
+      "step": 445
+    },
+    {
+      "epoch": 0.2508186441858845,
+      "grad_norm": 0.8662014603614807,
+      "learning_rate": 9.330937315754329e-05,
+      "loss": 1.5459,
+      "step": 450
+    },
+    {
+      "epoch": 0.2536055180101721,
+      "grad_norm": 0.8968704342842102,
+      "learning_rate": 9.306430768627753e-05,
+      "loss": 1.5309,
+      "step": 455
+    },
+    {
+      "epoch": 0.2563923918344597,
+      "grad_norm": 0.915246844291687,
+      "learning_rate": 9.281516833018571e-05,
+      "loss": 1.552,
+      "step": 460
+    },
+    {
+      "epoch": 0.2591792656587473,
+      "grad_norm": 0.913821280002594,
+      "learning_rate": 9.256197865785854e-05,
+      "loss": 1.6284,
+      "step": 465
+    },
+    {
+      "epoch": 0.26196613948303493,
+      "grad_norm": 0.8203092217445374,
+      "learning_rate": 9.230476262104677e-05,
+      "loss": 1.5235,
+      "step": 470
+    },
+    {
+      "epoch": 0.26475301330732254,
+      "grad_norm": 0.8390527963638306,
+      "learning_rate": 9.204354455239539e-05,
+      "loss": 1.5721,
+      "step": 475
+    },
+    {
+      "epoch": 0.26753988713161014,
+      "grad_norm": 0.8474882245063782,
+      "learning_rate": 9.177834916314165e-05,
+      "loss": 1.5674,
+      "step": 480
+    },
+    {
+      "epoch": 0.27032676095589775,
+      "grad_norm": 0.8627411127090454,
+      "learning_rate": 9.150920154077754e-05,
+      "loss": 1.5457,
+      "step": 485
+    },
+    {
+      "epoch": 0.27311363478018535,
+      "grad_norm": 0.9049586057662964,
+      "learning_rate": 9.123612714667634e-05,
+      "loss": 1.5827,
+      "step": 490
+    },
+    {
+      "epoch": 0.2759005086044729,
+      "grad_norm": 0.806458055973053,
+      "learning_rate": 9.095915181368412e-05,
+      "loss": 1.6398,
+      "step": 495
+    },
+    {
+      "epoch": 0.2786873824287605,
+      "grad_norm": 0.9326282143592834,
+      "learning_rate": 9.067830174367586e-05,
+      "loss": 1.6183,
+      "step": 500
+    },
+    {
+      "epoch": 0.2814742562530481,
+      "grad_norm": 0.8715500831604004,
+      "learning_rate": 9.039360350507679e-05,
+      "loss": 1.5529,
+      "step": 505
+    },
+    {
+      "epoch": 0.2842611300773357,
+      "grad_norm": 0.9053698778152466,
+      "learning_rate": 9.010508403034898e-05,
+      "loss": 1.5207,
+      "step": 510
+    },
+    {
+      "epoch": 0.2870480039016233,
+      "grad_norm": 0.8616856932640076,
+      "learning_rate": 8.98127706134436e-05,
+      "loss": 1.5696,
+      "step": 515
+    },
+    {
+      "epoch": 0.28983487772591093,
+      "grad_norm": 0.9130306243896484,
+      "learning_rate": 8.951669090721881e-05,
+      "loss": 1.5702,
+      "step": 520
+    },
+    {
+      "epoch": 0.29262175155019854,
+      "grad_norm": 0.8649139404296875,
+      "learning_rate": 8.921687292082393e-05,
+      "loss": 1.4826,
+      "step": 525
+    },
+    {
+      "epoch": 0.29540862537448614,
+      "grad_norm": 0.8023464679718018,
+      "learning_rate": 8.891334501704962e-05,
+      "loss": 1.5544,
+      "step": 530
+    },
+    {
+      "epoch": 0.29819549919877375,
+      "grad_norm": 0.9166476726531982,
+      "learning_rate": 8.86061359096449e-05,
+      "loss": 1.5016,
+      "step": 535
+    },
+    {
+      "epoch": 0.30098237302306136,
+      "grad_norm": 0.8918821811676025,
+      "learning_rate": 8.829527466060072e-05,
+      "loss": 1.416,
+      "step": 540
+    },
+    {
+      "epoch": 0.30376924684734896,
+      "grad_norm": 0.8663053512573242,
+      "learning_rate": 8.798079067740077e-05,
+      "loss": 1.641,
+      "step": 545
+    },
+    {
+      "epoch": 0.30655612067163657,
+      "grad_norm": 0.9275104403495789,
+      "learning_rate": 8.766271371023948e-05,
+      "loss": 1.5318,
+      "step": 550
+    },
+    {
+      "epoch": 0.3093429944959242,
+      "grad_norm": 0.9378035664558411,
+      "learning_rate": 8.73410738492077e-05,
+      "loss": 1.4886,
+      "step": 555
+    },
+    {
+      "epoch": 0.3121298683202118,
+      "grad_norm": 0.8419756889343262,
+      "learning_rate": 8.701590152144612e-05,
+      "loss": 1.5012,
+      "step": 560
+    },
+    {
+      "epoch": 0.3149167421444994,
+      "grad_norm": 0.8785234093666077,
+      "learning_rate": 8.668722748826693e-05,
+      "loss": 1.5278,
+      "step": 565
+    },
+    {
+      "epoch": 0.317703615968787,
+      "grad_norm": 0.9210188984870911,
+      "learning_rate": 8.635508284224371e-05,
+      "loss": 1.6802,
+      "step": 570
+    },
+    {
+      "epoch": 0.3204904897930746,
+      "grad_norm": 0.8633784651756287,
+      "learning_rate": 8.601949900427016e-05,
+      "loss": 1.5038,
+      "step": 575
+    },
+    {
+      "epoch": 0.3232773636173622,
+      "grad_norm": 0.9098255634307861,
+      "learning_rate": 8.568050772058762e-05,
+      "loss": 1.5611,
+      "step": 580
+    },
+    {
+      "epoch": 0.3260642374416498,
+      "grad_norm": 0.9021430015563965,
+      "learning_rate": 8.533814105978191e-05,
+      "loss": 1.5166,
+      "step": 585
+    },
+    {
+      "epoch": 0.3288511112659374,
+      "grad_norm": 0.8397455215454102,
+      "learning_rate": 8.499243140974966e-05,
+      "loss": 1.5493,
+      "step": 590
+    },
+    {
+      "epoch": 0.331637985090225,
+      "grad_norm": 0.8716102838516235,
+      "learning_rate": 8.464341147463431e-05,
+      "loss": 1.5242,
+      "step": 595
+    },
+    {
+      "epoch": 0.3344248589145126,
+      "grad_norm": 0.8432428240776062,
+      "learning_rate": 8.429111427173241e-05,
+      "loss": 1.4659,
+      "step": 600
+    },
+    {
+      "epoch": 0.33721173273880023,
+      "grad_norm": 0.8041474223136902,
+      "learning_rate": 8.393557312837018e-05,
+      "loss": 1.4995,
+      "step": 605
+    },
+    {
+      "epoch": 0.33999860656308784,
+      "grad_norm": 0.8596203327178955,
+      "learning_rate": 8.357682167875062e-05,
+      "loss": 1.5639,
+      "step": 610
+    },
+    {
+      "epoch": 0.34278548038737544,
+      "grad_norm": 0.888679027557373,
+      "learning_rate": 8.321489386077192e-05,
+      "loss": 1.5269,
+      "step": 615
+    },
+    {
+      "epoch": 0.34557235421166305,
+      "grad_norm": 0.9395999312400818,
+      "learning_rate": 8.28498239128167e-05,
+      "loss": 1.5474,
+      "step": 620
+    },
+    {
+      "epoch": 0.34835922803595065,
+      "grad_norm": 0.8320375084877014,
+      "learning_rate": 8.248164637051321e-05,
+      "loss": 1.4813,
+      "step": 625
+    },
+    {
+      "epoch": 0.35114610186023826,
+      "grad_norm": 0.901525616645813,
+      "learning_rate": 8.211039606346826e-05,
+      "loss": 1.5674,
+      "step": 630
+    },
+    {
+      "epoch": 0.35393297568452586,
+      "grad_norm": 0.8870389461517334,
+      "learning_rate": 8.173610811197226e-05,
+      "loss": 1.5355,
+      "step": 635
+    },
+    {
+      "epoch": 0.35671984950881347,
+      "grad_norm": 0.8773472309112549,
+      "learning_rate": 8.135881792367686e-05,
+      "loss": 1.5594,
+      "step": 640
+    },
+    {
+      "epoch": 0.3595067233331011,
+      "grad_norm": 0.8549400568008423,
+      "learning_rate": 8.097856119024545e-05,
+      "loss": 1.6111,
+      "step": 645
+    },
+    {
+      "epoch": 0.3622935971573887,
+      "grad_norm": 0.8216549754142761,
+      "learning_rate": 8.059537388397665e-05,
+      "loss": 1.5412,
+      "step": 650
+    },
+    {
+      "epoch": 0.3650804709816763,
+      "grad_norm": 0.8757383227348328,
+      "learning_rate": 8.020929225440137e-05,
+      "loss": 1.544,
+      "step": 655
+    },
+    {
+      "epoch": 0.3678673448059639,
+      "grad_norm": 0.8638432025909424,
+      "learning_rate": 7.98203528248536e-05,
+      "loss": 1.5394,
+      "step": 660
+    },
+    {
+      "epoch": 0.3706542186302515,
+      "grad_norm": 0.9091057181358337,
+      "learning_rate": 7.942859238901528e-05,
+      "loss": 1.5911,
+      "step": 665
+    },
+    {
+      "epoch": 0.3734410924545391,
+      "grad_norm": 1.0159931182861328,
+      "learning_rate": 7.903404800743564e-05,
+      "loss": 1.5694,
+      "step": 670
+    },
+    {
+      "epoch": 0.3762279662788267,
+      "grad_norm": 0.9712515473365784,
+      "learning_rate": 7.863675700402526e-05,
+      "loss": 1.5181,
+      "step": 675
+    },
+    {
+      "epoch": 0.3790148401031143,
+      "grad_norm": 0.9262729287147522,
+      "learning_rate": 7.823675696252524e-05,
+      "loss": 1.604,
+      "step": 680
+    },
+    {
+      "epoch": 0.3818017139274019,
+      "grad_norm": 0.8269544839859009,
+      "learning_rate": 7.783408572295174e-05,
+      "loss": 1.5848,
+      "step": 685
+    },
+    {
+      "epoch": 0.3845885877516895,
+      "grad_norm": 0.8753747344017029,
+      "learning_rate": 7.742878137801639e-05,
+      "loss": 1.5112,
+      "step": 690
+    },
+    {
+      "epoch": 0.38737546157597713,
+      "grad_norm": 0.834754228591919,
+      "learning_rate": 7.702088226952258e-05,
+      "loss": 1.4967,
+      "step": 695
+    },
+    {
+      "epoch": 0.39016233540026474,
+      "grad_norm": 0.9299247860908508,
+      "learning_rate": 7.661042698473853e-05,
+      "loss": 1.4628,
+      "step": 700
+    },
+    {
+      "epoch": 0.39294920922455234,
+      "grad_norm": 0.8204703330993652,
+      "learning_rate": 7.619745435274667e-05,
+      "loss": 1.493,
+      "step": 705
+    },
+    {
+      "epoch": 0.39573608304883995,
+      "grad_norm": 0.9027642011642456,
+      "learning_rate": 7.578200344077073e-05,
+      "loss": 1.5512,
+      "step": 710
+    },
+    {
+      "epoch": 0.39852295687312755,
+      "grad_norm": 0.923340380191803,
+      "learning_rate": 7.536411355047964e-05,
+      "loss": 1.5396,
+      "step": 715
+    },
+    {
+      "epoch": 0.40130983069741516,
+      "grad_norm": 0.855146586894989,
+      "learning_rate": 7.494382421426984e-05,
+      "loss": 1.5894,
+      "step": 720
+    },
+    {
+      "epoch": 0.40409670452170277,
+      "grad_norm": 0.850355327129364,
+      "learning_rate": 7.452117519152542e-05,
+      "loss": 1.474,
+      "step": 725
+    },
+    {
+      "epoch": 0.40688357834599037,
+      "grad_norm": 0.9606896638870239,
+      "learning_rate": 7.409620646485685e-05,
+      "loss": 1.506,
+      "step": 730
+    },
+    {
+      "epoch": 0.409670452170278,
+      "grad_norm": 0.8371616005897522,
+      "learning_rate": 7.36689582363187e-05,
+      "loss": 1.5454,
+      "step": 735
+    },
+    {
+      "epoch": 0.4124573259945656,
+      "grad_norm": 0.8478779792785645,
+      "learning_rate": 7.323947092360649e-05,
+      "loss": 1.4607,
+      "step": 740
+    },
+    {
+      "epoch": 0.4152441998188532,
+      "grad_norm": 0.9579894542694092,
+      "learning_rate": 7.280778515623314e-05,
+      "loss": 1.5948,
+      "step": 745
+    },
+    {
+      "epoch": 0.4180310736431408,
+      "grad_norm": 0.9187178015708923,
+      "learning_rate": 7.237394177168548e-05,
+      "loss": 1.5587,
+      "step": 750
+    },
+    {
+      "epoch": 0.4208179474674284,
+      "grad_norm": 0.9753735661506653,
+      "learning_rate": 7.193798181156095e-05,
+      "loss": 1.5907,
+      "step": 755
+    },
+    {
+      "epoch": 0.423604821291716,
+      "grad_norm": 0.8725233674049377,
+      "learning_rate": 7.149994651768514e-05,
+      "loss": 1.4151,
+      "step": 760
+    },
+    {
+      "epoch": 0.4263916951160036,
+      "grad_norm": 0.9570207595825195,
+      "learning_rate": 7.10598773282103e-05,
+      "loss": 1.5483,
+      "step": 765
+    },
+    {
+      "epoch": 0.4291785689402912,
+      "grad_norm": 0.9618322253227234,
+      "learning_rate": 7.061781587369519e-05,
+      "loss": 1.6014,
+      "step": 770
+    },
+    {
+      "epoch": 0.4319654427645788,
+      "grad_norm": 0.8392998576164246,
+      "learning_rate": 7.017380397316695e-05,
+      "loss": 1.4826,
+      "step": 775
+    },
+    {
+      "epoch": 0.4347523165888664,
+      "grad_norm": 0.8500150442123413,
+      "learning_rate": 6.972788363016497e-05,
+      "loss": 1.4199,
+      "step": 780
+    },
+    {
+      "epoch": 0.43753919041315403,
+      "grad_norm": 0.9263842105865479,
+      "learning_rate": 6.92800970287674e-05,
+      "loss": 1.5328,
+      "step": 785
+    },
+    {
+      "epoch": 0.44032606423744164,
+      "grad_norm": 0.8920561671257019,
+      "learning_rate": 6.883048652960038e-05,
+      "loss": 1.5331,
+      "step": 790
+    },
+    {
+      "epoch": 0.44311293806172924,
+      "grad_norm": 0.8388264775276184,
+      "learning_rate": 6.837909466583095e-05,
+      "loss": 1.5442,
+      "step": 795
+    },
+    {
+      "epoch": 0.44589981188601685,
+      "grad_norm": 0.8242114186286926,
+      "learning_rate": 6.792596413914324e-05,
+      "loss": 1.4163,
+      "step": 800
+    },
+    {
+      "epoch": 0.44868668571030446,
+      "grad_norm": 0.9328235387802124,
+      "learning_rate": 6.747113781569892e-05,
+      "loss": 1.5607,
+      "step": 805
+    },
+    {
+      "epoch": 0.45147355953459206,
+      "grad_norm": 0.9171831607818604,
+      "learning_rate": 6.701465872208216e-05,
+      "loss": 1.5812,
+      "step": 810
+    },
+    {
+      "epoch": 0.45426043335887967,
+      "grad_norm": 0.9111394882202148,
+      "learning_rate": 6.655657004122916e-05,
+      "loss": 1.5764,
+      "step": 815
+    },
+    {
+      "epoch": 0.4570473071831673,
+      "grad_norm": 0.9669636487960815,
+      "learning_rate": 6.60969151083432e-05,
+      "loss": 1.5125,
+      "step": 820
+    },
+    {
+      "epoch": 0.4598341810074549,
+      "grad_norm": 0.8893470168113708,
+      "learning_rate": 6.563573740679496e-05,
+      "loss": 1.5136,
+      "step": 825
+    },
+    {
+      "epoch": 0.4626210548317425,
+      "grad_norm": 0.8966957926750183,
+      "learning_rate": 6.517308056400917e-05,
+      "loss": 1.5365,
+      "step": 830
+    },
+    {
+      "epoch": 0.4654079286560301,
+      "grad_norm": 0.9211333990097046,
+      "learning_rate": 6.470898834733731e-05,
+      "loss": 1.5054,
+      "step": 835
+    },
+    {
+      "epoch": 0.4681948024803177,
+      "grad_norm": 0.8702636361122131,
+      "learning_rate": 6.42435046599173e-05,
+      "loss": 1.5078,
+      "step": 840
+    },
+    {
+      "epoch": 0.4709816763046053,
+      "grad_norm": 1.0185481309890747,
+      "learning_rate": 6.377667353652022e-05,
+      "loss": 1.5513,
+      "step": 845
+    },
+    {
+      "epoch": 0.4737685501288929,
+      "grad_norm": 0.8714690208435059,
+      "learning_rate": 6.330853913938466e-05,
+      "loss": 1.4742,
+      "step": 850
+    },
+    {
+      "epoch": 0.4765554239531805,
+      "grad_norm": 0.8962568044662476,
+      "learning_rate": 6.283914575403888e-05,
+      "loss": 1.5775,
+      "step": 855
+    },
+    {
+      "epoch": 0.4793422977774681,
+      "grad_norm": 0.9320731163024902,
+      "learning_rate": 6.236853778511156e-05,
+      "loss": 1.5052,
+      "step": 860
+    },
+    {
+      "epoch": 0.4821291716017557,
+      "grad_norm": 0.9507172703742981,
+      "learning_rate": 6.189675975213094e-05,
+      "loss": 1.5255,
+      "step": 865
+    },
+    {
+      "epoch": 0.48491604542604333,
+      "grad_norm": 0.9428796172142029,
+      "learning_rate": 6.142385628531342e-05,
+      "loss": 1.5391,
+      "step": 870
+    },
+    {
+      "epoch": 0.48770291925033094,
+      "grad_norm": 0.9637115001678467,
+      "learning_rate": 6.09498721213414e-05,
+      "loss": 1.4856,
+      "step": 875
+    },
+    {
+      "epoch": 0.49048979307461854,
+      "grad_norm": 0.9000864624977112,
+      "learning_rate": 6.047485209913137e-05,
+      "loss": 1.6011,
+      "step": 880
+    },
+    {
+      "epoch": 0.49327666689890615,
+      "grad_norm": 0.820088803768158,
+      "learning_rate": 5.999884115559192e-05,
+      "loss": 1.4913,
+      "step": 885
+    },
+    {
+      "epoch": 0.49606354072319375,
+      "grad_norm": 0.7902366518974304,
+      "learning_rate": 5.952188432137293e-05,
+      "loss": 1.5304,
+      "step": 890
+    },
+    {
+      "epoch": 0.49885041454748136,
+      "grad_norm": 0.7885243892669678,
+      "learning_rate": 5.90440267166055e-05,
+      "loss": 1.4527,
+      "step": 895
+    },
+    {
+      "epoch": 0.501637288371769,
+      "grad_norm": 1.0124362707138062,
+      "learning_rate": 5.8565313546633684e-05,
+      "loss": 1.5122,
+      "step": 900
+    },
+    {
+      "epoch": 0.5044241621960566,
+      "grad_norm": 0.8640070557594299,
+      "learning_rate": 5.8085790097738025e-05,
+      "loss": 1.5547,
+      "step": 905
+    },
+    {
+      "epoch": 0.5072110360203442,
+      "grad_norm": 0.9227625131607056,
+      "learning_rate": 5.7605501732851475e-05,
+      "loss": 1.5804,
+      "step": 910
+    },
+    {
+      "epoch": 0.5099979098446318,
+      "grad_norm": 0.8405686616897583,
+      "learning_rate": 5.712449388726807e-05,
+      "loss": 1.5661,
+      "step": 915
+    },
+    {
+      "epoch": 0.5127847836689194,
+      "grad_norm": 0.867641031742096,
+      "learning_rate": 5.664281206434472e-05,
+      "loss": 1.4674,
+      "step": 920
+    },
+    {
+      "epoch": 0.515571657493207,
+      "grad_norm": 0.8705233931541443,
+      "learning_rate": 5.616050183119663e-05,
+      "loss": 1.5702,
+      "step": 925
+    },
+    {
+      "epoch": 0.5183585313174947,
+      "grad_norm": 0.8183286786079407,
+      "learning_rate": 5.5677608814386616e-05,
+      "loss": 1.3906,
+      "step": 930
+    },
+    {
+      "epoch": 0.5211454051417822,
+      "grad_norm": 0.9858644604682922,
+      "learning_rate": 5.519417869560889e-05,
+      "loss": 1.5129,
+      "step": 935
+    },
+    {
+      "epoch": 0.5239322789660699,
+      "grad_norm": 0.9389538168907166,
+      "learning_rate": 5.471025720736747e-05,
+      "loss": 1.55,
+      "step": 940
+    },
+    {
+      "epoch": 0.5267191527903574,
+      "grad_norm": 0.8737983107566833,
+      "learning_rate": 5.422589012864996e-05,
+      "loss": 1.5582,
+      "step": 945
+    },
+    {
+      "epoch": 0.5295060266146451,
+      "grad_norm": 0.9484609961509705,
+      "learning_rate": 5.3741123280596864e-05,
+      "loss": 1.4895,
+      "step": 950
+    },
+    {
+      "epoch": 0.5322929004389326,
+      "grad_norm": 0.9373424649238586,
+      "learning_rate": 5.325600252216685e-05,
+      "loss": 1.502,
+      "step": 955
+    },
+    {
+      "epoch": 0.5350797742632203,
+      "grad_norm": 0.9036156535148621,
+      "learning_rate": 5.27705737457985e-05,
+      "loss": 1.394,
+      "step": 960
+    },
+    {
+      "epoch": 0.5378666480875078,
+      "grad_norm": 0.9241552948951721,
+      "learning_rate": 5.228488287306896e-05,
+      "loss": 1.5101,
+      "step": 965
+    },
+    {
+      "epoch": 0.5406535219117955,
+      "grad_norm": 0.8161901235580444,
+      "learning_rate": 5.179897585034963e-05,
+      "loss": 1.5254,
+      "step": 970
+    },
+    {
+      "epoch": 0.543440395736083,
+      "grad_norm": 0.8650552034378052,
+      "learning_rate": 5.1312898644459776e-05,
+      "loss": 1.4365,
+      "step": 975
+    },
+    {
+      "epoch": 0.5462272695603707,
+      "grad_norm": 0.9877922534942627,
+      "learning_rate": 5.0826697238317935e-05,
+      "loss": 1.5132,
+      "step": 980
+    },
+    {
+      "epoch": 0.5490141433846583,
+      "grad_norm": 0.8969970941543579,
+      "learning_rate": 5.0340417626592016e-05,
+      "loss": 1.551,
+      "step": 985
+    },
+    {
+      "epoch": 0.5518010172089458,
+      "grad_norm": 0.9135898351669312,
+      "learning_rate": 4.9854105811348216e-05,
+      "loss": 1.5476,
+      "step": 990
+    },
+    {
+      "epoch": 0.5545878910332335,
+      "grad_norm": 0.9921842217445374,
+      "learning_rate": 4.936780779769913e-05,
+      "loss": 1.5394,
+      "step": 995
+    },
+    {
+      "epoch": 0.557374764857521,
+      "grad_norm": 0.8833547830581665,
+      "learning_rate": 4.888156958945174e-05,
+      "loss": 1.5002,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 1795,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.3496399537522048e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79ce14c8b9b5fd843768a164624a4f08d081472da21dd3d1eb9a404fc8d77b0d
+size 6161

checkpoint-1500/README.md

@@ -0,0 +1,202 @@
+---
+base_model: google/gemma-3-12b-it
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.15.2

checkpoint-1500/adapter_config.json

@@ -0,0 +1,180 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-3-12b-it",
+  "bias": "none",
+  "corda_config": null,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "language_model.layers.3.self_attn.q_proj",
+    "35.self_attn.k_proj",
+    "o_proj",
+    "40.self_attn.q_proj",
+    "language_model.layers.8.self_attn.k_proj",
+    "language_model.layers.25.self_attn.v_proj",
+    "language_model.layers.11.self_attn.k_proj",
+    "43.self_attn.q_proj",
+    "language_model.layers.15.self_attn.q_proj",
+    "39.self_attn.q_proj",
+    "language_model.layers.21.self_attn.v_proj",
+    "35.self_attn.q_proj",
+    "31.self_attn.v_proj",
+    "language_model.layers.8.self_attn.v_proj",
+    "38.self_attn.k_proj",
+    "language_model.layers.0.self_attn.k_proj",
+    "language_model.layers.12.self_attn.v_proj",
+    "27.self_attn.q_proj",
+    "language_model.layers.5.self_attn.q_proj",
+    "language_model.layers.2.self_attn.q_proj",
+    "language_model.layers.20.self_attn.k_proj",
+    "up_proj",
+    "language_model.layers.14.self_attn.q_proj",
+    "language_model.layers.3.self_attn.v_proj",
+    "language_model.layers.21.self_attn.q_proj",
+    "language_model.layers.26.self_attn.q_proj",
+    "down_proj",
+    "language_model.layers.23.self_attn.q_proj",
+    "language_model.layers.8.self_attn.q_proj",
+    "30.self_attn.v_proj",
+    "language_model.layers.1.self_attn.v_proj",
+    "36.self_attn.q_proj",
+    "28.self_attn.v_proj",
+    "language_model.layers.23.self_attn.v_proj",
+    "language_model.layers.22.self_attn.q_proj",
+    "language_model.layers.1.self_attn.k_proj",
+    "38.self_attn.q_proj",
+    "32.self_attn.q_proj",
+    "46.self_attn.k_proj",
+    "34.self_attn.k_proj",
+    "language_model.layers.16.self_attn.v_proj",
+    "language_model.layers.5.self_attn.v_proj",
+    "47.self_attn.k_proj",
+    "42.self_attn.k_proj",
+    "45.self_attn.k_proj",
+    "45.self_attn.v_proj",
+    "language_model.layers.19.self_attn.k_proj",
+    "language_model.layers.22.self_attn.k_proj",
+    "language_model.layers.17.self_attn.v_proj",
+    "language_model.layers.17.self_attn.q_proj",
+    "language_model.layers.9.self_attn.k_proj",
+    "language_model.layers.6.self_attn.q_proj",
+    "language_model.layers.1.self_attn.q_proj",
+    "language_model.layers.7.self_attn.k_proj",
+    "39.self_attn.k_proj",
+    "language_model.layers.6.self_attn.k_proj",
+    "language_model.layers.18.self_attn.q_proj",
+    "language_model.layers.5.self_attn.k_proj",
+    "29.self_attn.k_proj",
+    "language_model.layers.20.self_attn.v_proj",
+    "language_model.layers.25.self_attn.k_proj",
+    "language_model.layers.4.self_attn.q_proj",
+    "language_model.layers.13.self_attn.k_proj",
+    "41.self_attn.v_proj",
+    "language_model.layers.12.self_attn.q_proj",
+    "language_model.layers.7.self_attn.q_proj",
+    "language_model.layers.19.self_attn.v_proj",
+    "language_model.layers.7.self_attn.v_proj",
+    "language_model.layers.11.self_attn.q_proj",
+    "30.self_attn.k_proj",
+    "language_model.layers.16.self_attn.k_proj",
+    "43.self_attn.k_proj",
+    "47.self_attn.q_proj",
+    "language_model.layers.10.self_attn.v_proj",
+    "language_model.layers.15.self_attn.v_proj",
+    "language_model.layers.0.self_attn.v_proj",
+    "31.self_attn.k_proj",
+    "35.self_attn.v_proj",
+    "language_model.layers.2.self_attn.k_proj",
+    "27.self_attn.v_proj",
+    "30.self_attn.q_proj",
+    "28.self_attn.k_proj",
+    "41.self_attn.k_proj",
+    "language_model.layers.0.self_attn.q_proj",
+    "33.self_attn.v_proj",
+    "language_model.layers.6.self_attn.v_proj",
+    "44.self_attn.q_proj",
+    "language_model.layers.2.self_attn.v_proj",
+    "language_model.layers.3.self_attn.k_proj",
+    "language_model.layers.9.self_attn.q_proj",
+    "37.self_attn.q_proj",
+    "language_model.layers.4.self_attn.v_proj",
+    "33.self_attn.q_proj",
+    "29.self_attn.v_proj",
+    "language_model.layers.18.self_attn.k_proj",
+    "37.self_attn.v_proj",
+    "language_model.layers.17.self_attn.k_proj",
+    "33.self_attn.k_proj",
+    "language_model.layers.19.self_attn.q_proj",
+    "gate_proj",
+    "language_model.layers.10.self_attn.q_proj",
+    "47.self_attn.v_proj",
+    "42.self_attn.q_proj",
+    "language_model.layers.16.self_attn.q_proj",
+    "language_model.layers.25.self_attn.q_proj",
+    "42.self_attn.v_proj",
+    "32.self_attn.k_proj",
+    "language_model.layers.9.self_attn.v_proj",
+    "46.self_attn.v_proj",
+    "language_model.layers.12.self_attn.k_proj",
+    "38.self_attn.v_proj",
+    "language_model.layers.14.self_attn.k_proj",
+    "46.self_attn.q_proj",
+    "language_model.layers.18.self_attn.v_proj",
+    "language_model.layers.26.self_attn.v_proj",
+    "44.self_attn.v_proj",
+    "language_model.layers.24.self_attn.q_proj",
+    "language_model.layers.15.self_attn.k_proj",
+    "language_model.layers.14.self_attn.v_proj",
+    "29.self_attn.q_proj",
+    "language_model.layers.4.self_attn.k_proj",
+    "language_model.layers.13.self_attn.q_proj",
+    "43.self_attn.v_proj",
+    "34.self_attn.q_proj",
+    "45.self_attn.q_proj",
+    "32.self_attn.v_proj",
+    "language_model.layers.11.self_attn.v_proj",
+    "36.self_attn.v_proj",
+    "language_model.layers.21.self_attn.k_proj",
+    "31.self_attn.q_proj",
+    "37.self_attn.k_proj",
+    "39.self_attn.v_proj",
+    "language_model.layers.26.self_attn.k_proj",
+    "40.self_attn.k_proj",
+    "36.self_attn.k_proj",
+    "41.self_attn.q_proj",
+    "language_model.layers.24.self_attn.v_proj",
+    "language_model.layers.23.self_attn.k_proj",
+    "language_model.layers.10.self_attn.k_proj",
+    "27.self_attn.k_proj",
+    "language_model.layers.13.self_attn.v_proj",
+    "34.self_attn.v_proj",
+    "40.self_attn.v_proj",
+    "language_model.layers.20.self_attn.q_proj",
+    "language_model.layers.24.self_attn.k_proj",
+    "44.self_attn.k_proj",
+    "28.self_attn.q_proj",
+    "language_model.layers.22.self_attn.v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-1500/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b116eb6fca53641af6b393df8aff14e3f040ca6236868ad53cd475020d1f85e
+size 131040264

checkpoint-1500/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<image_soft_token>": 262144
+}

checkpoint-1500/chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

checkpoint-1500/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa2d9d9aada4ee775490ca548b6a64fc0af4c2bf9f09137c62e99092aa17aa44
+size 262448707

checkpoint-1500/preprocessor_config.json

@@ -0,0 +1,29 @@
+{
+  "do_convert_rgb": null,
+  "do_normalize": true,
+  "do_pan_and_scan": null,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "Gemma3ImageProcessor",
+  "image_seq_length": 256,
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "pan_and_scan_max_num_crops": null,
+  "pan_and_scan_min_crop_size": null,
+  "pan_and_scan_min_ratio_to_activate": null,
+  "processor_class": "Gemma3Processor",
+  "resample": 2,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "height": 896,
+    "width": 896
+  }
+}

checkpoint-1500/processor_config.json

@@ -0,0 +1,4 @@
+{
+  "image_seq_length": 256,
+  "processor_class": "Gemma3Processor"
+}

checkpoint-1500/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:250560ab3d528161ab3659b120def6e4a9ab4b457e3399603bbcfa40db3efc90
+size 14645

checkpoint-1500/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74ecf9f08b8ece20f8e38217288ba9577a3fcfff91bb28828bd121972f45e8e0
+size 1465

checkpoint-1500/special_tokens_map.json

@@ -0,0 +1,33 @@
+{
+  "boi_token": "<start_of_image>",
+  "bos_token": {
+    "content": "<bos>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eoi_token": "<end_of_image>",
+  "eos_token": {
+    "content": "<end_of_turn>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "image_token": "<image_soft_token>",
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-1500/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4667f2089529e8e7657cfb6d1c19910ae71ff5f28aa7ab2ff2763330affad795
+size 33384568

checkpoint-1500/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c
+size 4689074

checkpoint-1500/trainer_state.json

@@ -0,0 +1,2134 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8360621472862816,
+  "eval_steps": 500,
+  "global_step": 1500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0027868738242876052,
+      "grad_norm": 2.1914708614349365,
+      "learning_rate": 2.2222222222222225e-06,
+      "loss": 2.8194,
+      "step": 5
+    },
+    {
+      "epoch": 0.0055737476485752105,
+      "grad_norm": 2.056274890899658,
+      "learning_rate": 5e-06,
+      "loss": 2.7461,
+      "step": 10
+    },
+    {
+      "epoch": 0.008360621472862817,
+      "grad_norm": 1.9419809579849243,
+      "learning_rate": 7.777777777777777e-06,
+      "loss": 2.7894,
+      "step": 15
+    },
+    {
+      "epoch": 0.011147495297150421,
+      "grad_norm": 2.0836448669433594,
+      "learning_rate": 1.0555555555555555e-05,
+      "loss": 2.8633,
+      "step": 20
+    },
+    {
+      "epoch": 0.013934369121438027,
+      "grad_norm": 2.413074493408203,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 2.7385,
+      "step": 25
+    },
+    {
+      "epoch": 0.016721242945725634,
+      "grad_norm": 2.3652689456939697,
+      "learning_rate": 1.6111111111111115e-05,
+      "loss": 2.6383,
+      "step": 30
+    },
+    {
+      "epoch": 0.019508116770013236,
+      "grad_norm": 1.4635894298553467,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 2.3059,
+      "step": 35
+    },
+    {
+      "epoch": 0.022294990594300842,
+      "grad_norm": 1.319265365600586,
+      "learning_rate": 2.1666666666666667e-05,
+      "loss": 2.0529,
+      "step": 40
+    },
+    {
+      "epoch": 0.025081864418588447,
+      "grad_norm": 1.2513976097106934,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 2.0179,
+      "step": 45
+    },
+    {
+      "epoch": 0.027868738242876053,
+      "grad_norm": 1.264099359512329,
+      "learning_rate": 2.7222222222222223e-05,
+      "loss": 1.9416,
+      "step": 50
+    },
+    {
+      "epoch": 0.03065561206716366,
+      "grad_norm": 1.0929852724075317,
+      "learning_rate": 3e-05,
+      "loss": 1.7526,
+      "step": 55
+    },
+    {
+      "epoch": 0.03344248589145127,
+      "grad_norm": 1.0694290399551392,
+      "learning_rate": 3.277777777777778e-05,
+      "loss": 1.9019,
+      "step": 60
+    },
+    {
+      "epoch": 0.03622935971573887,
+      "grad_norm": 1.2505921125411987,
+      "learning_rate": 3.555555555555556e-05,
+      "loss": 1.7937,
+      "step": 65
+    },
+    {
+      "epoch": 0.03901623354002647,
+      "grad_norm": 1.1209263801574707,
+      "learning_rate": 3.8333333333333334e-05,
+      "loss": 1.6891,
+      "step": 70
+    },
+    {
+      "epoch": 0.04180310736431408,
+      "grad_norm": 1.1830799579620361,
+      "learning_rate": 4.111111111111111e-05,
+      "loss": 1.7494,
+      "step": 75
+    },
+    {
+      "epoch": 0.044589981188601684,
+      "grad_norm": 1.0991789102554321,
+      "learning_rate": 4.388888888888889e-05,
+      "loss": 1.6545,
+      "step": 80
+    },
+    {
+      "epoch": 0.04737685501288929,
+      "grad_norm": 1.075130820274353,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 1.7803,
+      "step": 85
+    },
+    {
+      "epoch": 0.050163728837176895,
+      "grad_norm": 1.1547001600265503,
+      "learning_rate": 4.9444444444444446e-05,
+      "loss": 1.6842,
+      "step": 90
+    },
+    {
+      "epoch": 0.0529506026614645,
+      "grad_norm": 1.2314906120300293,
+      "learning_rate": 5.222222222222223e-05,
+      "loss": 1.6707,
+      "step": 95
+    },
+    {
+      "epoch": 0.055737476485752106,
+      "grad_norm": 1.1287559270858765,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 1.6046,
+      "step": 100
+    },
+    {
+      "epoch": 0.05852435031003971,
+      "grad_norm": 1.0470261573791504,
+      "learning_rate": 5.7777777777777776e-05,
+      "loss": 1.6453,
+      "step": 105
+    },
+    {
+      "epoch": 0.06131122413432732,
+      "grad_norm": 1.3171180486679077,
+      "learning_rate": 6.055555555555555e-05,
+      "loss": 1.7238,
+      "step": 110
+    },
+    {
+      "epoch": 0.06409809795861493,
+      "grad_norm": 1.1685153245925903,
+      "learning_rate": 6.333333333333333e-05,
+      "loss": 1.6259,
+      "step": 115
+    },
+    {
+      "epoch": 0.06688497178290254,
+      "grad_norm": 1.006691813468933,
+      "learning_rate": 6.611111111111111e-05,
+      "loss": 1.6396,
+      "step": 120
+    },
+    {
+      "epoch": 0.06967184560719013,
+      "grad_norm": 1.1026668548583984,
+      "learning_rate": 6.88888888888889e-05,
+      "loss": 1.7253,
+      "step": 125
+    },
+    {
+      "epoch": 0.07245871943147773,
+      "grad_norm": 1.1105059385299683,
+      "learning_rate": 7.166666666666667e-05,
+      "loss": 1.6867,
+      "step": 130
+    },
+    {
+      "epoch": 0.07524559325576534,
+      "grad_norm": 1.0630534887313843,
+      "learning_rate": 7.444444444444444e-05,
+      "loss": 1.6371,
+      "step": 135
+    },
+    {
+      "epoch": 0.07803246708005294,
+      "grad_norm": 1.0416017770767212,
+      "learning_rate": 7.722222222222223e-05,
+      "loss": 1.643,
+      "step": 140
+    },
+    {
+      "epoch": 0.08081934090434055,
+      "grad_norm": 0.9604218006134033,
+      "learning_rate": 8e-05,
+      "loss": 1.7327,
+      "step": 145
+    },
+    {
+      "epoch": 0.08360621472862816,
+      "grad_norm": 1.033799648284912,
+      "learning_rate": 8.277777777777778e-05,
+      "loss": 1.6771,
+      "step": 150
+    },
+    {
+      "epoch": 0.08639308855291576,
+      "grad_norm": 0.9786093831062317,
+      "learning_rate": 8.555555555555556e-05,
+      "loss": 1.6525,
+      "step": 155
+    },
+    {
+      "epoch": 0.08917996237720337,
+      "grad_norm": 1.0249221324920654,
+      "learning_rate": 8.833333333333333e-05,
+      "loss": 1.6163,
+      "step": 160
+    },
+    {
+      "epoch": 0.09196683620149097,
+      "grad_norm": 1.0517358779907227,
+      "learning_rate": 9.111111111111112e-05,
+      "loss": 1.5926,
+      "step": 165
+    },
+    {
+      "epoch": 0.09475371002577858,
+      "grad_norm": 1.0581194162368774,
+      "learning_rate": 9.388888888888889e-05,
+      "loss": 1.6449,
+      "step": 170
+    },
+    {
+      "epoch": 0.09754058385006618,
+      "grad_norm": 1.0320991277694702,
+      "learning_rate": 9.666666666666667e-05,
+      "loss": 1.6547,
+      "step": 175
+    },
+    {
+      "epoch": 0.10032745767435379,
+      "grad_norm": 1.012792706489563,
+      "learning_rate": 9.944444444444446e-05,
+      "loss": 1.6617,
+      "step": 180
+    },
+    {
+      "epoch": 0.1031143314986414,
+      "grad_norm": 1.0023682117462158,
+      "learning_rate": 9.999848639521432e-05,
+      "loss": 1.598,
+      "step": 185
+    },
+    {
+      "epoch": 0.105901205322929,
+      "grad_norm": 1.0282474756240845,
+      "learning_rate": 9.999233753283091e-05,
+      "loss": 1.5899,
+      "step": 190
+    },
+    {
+      "epoch": 0.1086880791472166,
+      "grad_norm": 1.2141717672348022,
+      "learning_rate": 9.998145939378577e-05,
+      "loss": 1.585,
+      "step": 195
+    },
+    {
+      "epoch": 0.11147495297150421,
+      "grad_norm": 0.9987207651138306,
+      "learning_rate": 9.996585300715116e-05,
+      "loss": 1.6266,
+      "step": 200
+    },
+    {
+      "epoch": 0.11426182679579182,
+      "grad_norm": 0.9829638600349426,
+      "learning_rate": 9.994551984929175e-05,
+      "loss": 1.6098,
+      "step": 205
+    },
+    {
+      "epoch": 0.11704870062007942,
+      "grad_norm": 0.8391382098197937,
+      "learning_rate": 9.992046184372492e-05,
+      "loss": 1.628,
+      "step": 210
+    },
+    {
+      "epoch": 0.11983557444436703,
+      "grad_norm": 0.8903898596763611,
+      "learning_rate": 9.989068136093873e-05,
+      "loss": 1.5645,
+      "step": 215
+    },
+    {
+      "epoch": 0.12262244826865464,
+      "grad_norm": 0.9336121082305908,
+      "learning_rate": 9.985618121816779e-05,
+      "loss": 1.6621,
+      "step": 220
+    },
+    {
+      "epoch": 0.12540932209294225,
+      "grad_norm": 0.9730141162872314,
+      "learning_rate": 9.981696467912664e-05,
+      "loss": 1.5838,
+      "step": 225
+    },
+    {
+      "epoch": 0.12819619591722986,
+      "grad_norm": 0.9329681992530823,
+      "learning_rate": 9.97730354537011e-05,
+      "loss": 1.6099,
+      "step": 230
+    },
+    {
+      "epoch": 0.13098306974151747,
+      "grad_norm": 0.8763962984085083,
+      "learning_rate": 9.972439769759722e-05,
+      "loss": 1.516,
+      "step": 235
+    },
+    {
+      "epoch": 0.13376994356580507,
+      "grad_norm": 0.9208707213401794,
+      "learning_rate": 9.967105601194823e-05,
+      "loss": 1.6798,
+      "step": 240
+    },
+    {
+      "epoch": 0.13655681739009268,
+      "grad_norm": 0.9264675378799438,
+      "learning_rate": 9.961301544287922e-05,
+      "loss": 1.545,
+      "step": 245
+    },
+    {
+      "epoch": 0.13934369121438026,
+      "grad_norm": 1.0418734550476074,
+      "learning_rate": 9.955028148102979e-05,
+      "loss": 1.5595,
+      "step": 250
+    },
+    {
+      "epoch": 0.14213056503866786,
+      "grad_norm": 0.9690463542938232,
+      "learning_rate": 9.948286006103466e-05,
+      "loss": 1.6868,
+      "step": 255
+    },
+    {
+      "epoch": 0.14491743886295547,
+      "grad_norm": 0.9389734864234924,
+      "learning_rate": 9.941075756096226e-05,
+      "loss": 1.6089,
+      "step": 260
+    },
+    {
+      "epoch": 0.14770431268724307,
+      "grad_norm": 0.902563750743866,
+      "learning_rate": 9.933398080171123e-05,
+      "loss": 1.6093,
+      "step": 265
+    },
+    {
+      "epoch": 0.15049118651153068,
+      "grad_norm": 0.9173883199691772,
+      "learning_rate": 9.925253704636543e-05,
+      "loss": 1.5829,
+      "step": 270
+    },
+    {
+      "epoch": 0.15327806033581828,
+      "grad_norm": 0.9028013944625854,
+      "learning_rate": 9.916643399950656e-05,
+      "loss": 1.5999,
+      "step": 275
+    },
+    {
+      "epoch": 0.1560649341601059,
+      "grad_norm": 0.9312839508056641,
+      "learning_rate": 9.907567980648549e-05,
+      "loss": 1.5684,
+      "step": 280
+    },
+    {
+      "epoch": 0.1588518079843935,
+      "grad_norm": 0.8616472482681274,
+      "learning_rate": 9.898028305265169e-05,
+      "loss": 1.6562,
+      "step": 285
+    },
+    {
+      "epoch": 0.1616386818086811,
+      "grad_norm": 0.8607208728790283,
+      "learning_rate": 9.888025276254096e-05,
+      "loss": 1.6552,
+      "step": 290
+    },
+    {
+      "epoch": 0.1644255556329687,
+      "grad_norm": 0.8621200323104858,
+      "learning_rate": 9.877559839902184e-05,
+      "loss": 1.5795,
+      "step": 295
+    },
+    {
+      "epoch": 0.1672124294572563,
+      "grad_norm": 0.9334512948989868,
+      "learning_rate": 9.86663298624003e-05,
+      "loss": 1.5635,
+      "step": 300
+    },
+    {
+      "epoch": 0.16999930328154392,
+      "grad_norm": 0.9203838109970093,
+      "learning_rate": 9.855245748948326e-05,
+      "loss": 1.6554,
+      "step": 305
+    },
+    {
+      "epoch": 0.17278617710583152,
+      "grad_norm": 0.9085136651992798,
+      "learning_rate": 9.843399205260068e-05,
+      "loss": 1.6244,
+      "step": 310
+    },
+    {
+      "epoch": 0.17557305093011913,
+      "grad_norm": 1.005462884902954,
+      "learning_rate": 9.831094475858652e-05,
+      "loss": 1.6093,
+      "step": 315
+    },
+    {
+      "epoch": 0.17835992475440673,
+      "grad_norm": 0.9284034371376038,
+      "learning_rate": 9.818332724771857e-05,
+      "loss": 1.6564,
+      "step": 320
+    },
+    {
+      "epoch": 0.18114679857869434,
+      "grad_norm": 0.8924919962882996,
+      "learning_rate": 9.805115159261726e-05,
+      "loss": 1.5466,
+      "step": 325
+    },
+    {
+      "epoch": 0.18393367240298195,
+      "grad_norm": 0.8804652094841003,
+      "learning_rate": 9.791443029710361e-05,
+      "loss": 1.5292,
+      "step": 330
+    },
+    {
+      "epoch": 0.18672054622726955,
+      "grad_norm": 0.9378550052642822,
+      "learning_rate": 9.777317629501636e-05,
+      "loss": 1.5443,
+      "step": 335
+    },
+    {
+      "epoch": 0.18950742005155716,
+      "grad_norm": 0.812350869178772,
+      "learning_rate": 9.762740294898846e-05,
+      "loss": 1.5034,
+      "step": 340
+    },
+    {
+      "epoch": 0.19229429387584476,
+      "grad_norm": 0.871232271194458,
+      "learning_rate": 9.747712404918286e-05,
+      "loss": 1.4828,
+      "step": 345
+    },
+    {
+      "epoch": 0.19508116770013237,
+      "grad_norm": 1.0012617111206055,
+      "learning_rate": 9.732235381198813e-05,
+      "loss": 1.6139,
+      "step": 350
+    },
+    {
+      "epoch": 0.19786804152441997,
+      "grad_norm": 0.8306112289428711,
+      "learning_rate": 9.716310687867342e-05,
+      "loss": 1.4516,
+      "step": 355
+    },
+    {
+      "epoch": 0.20065491534870758,
+      "grad_norm": 0.925584614276886,
+      "learning_rate": 9.699939831400351e-05,
+      "loss": 1.6499,
+      "step": 360
+    },
+    {
+      "epoch": 0.20344178917299519,
+      "grad_norm": 0.9054883718490601,
+      "learning_rate": 9.683124360481364e-05,
+      "loss": 1.6033,
+      "step": 365
+    },
+    {
+      "epoch": 0.2062286629972828,
+      "grad_norm": 0.9771924614906311,
+      "learning_rate": 9.665865865854445e-05,
+      "loss": 1.5691,
+      "step": 370
+    },
+    {
+      "epoch": 0.2090155368215704,
+      "grad_norm": 0.9042426347732544,
+      "learning_rate": 9.648165980173712e-05,
+      "loss": 1.5722,
+      "step": 375
+    },
+    {
+      "epoch": 0.211802410645858,
+      "grad_norm": 0.9686840772628784,
+      "learning_rate": 9.630026377848892e-05,
+      "loss": 1.5936,
+      "step": 380
+    },
+    {
+      "epoch": 0.2145892844701456,
+      "grad_norm": 0.8492907285690308,
+      "learning_rate": 9.611448774886924e-05,
+      "loss": 1.5515,
+      "step": 385
+    },
+    {
+      "epoch": 0.2173761582944332,
+      "grad_norm": 0.95368891954422,
+      "learning_rate": 9.592434928729616e-05,
+      "loss": 1.6242,
+      "step": 390
+    },
+    {
+      "epoch": 0.22016303211872082,
+      "grad_norm": 0.9186941981315613,
+      "learning_rate": 9.572986638087396e-05,
+      "loss": 1.5976,
+      "step": 395
+    },
+    {
+      "epoch": 0.22294990594300843,
+      "grad_norm": 0.8386760354042053,
+      "learning_rate": 9.553105742769154e-05,
+      "loss": 1.5469,
+      "step": 400
+    },
+    {
+      "epoch": 0.22573677976729603,
+      "grad_norm": 0.875751793384552,
+      "learning_rate": 9.532794123508197e-05,
+      "loss": 1.4213,
+      "step": 405
+    },
+    {
+      "epoch": 0.22852365359158364,
+      "grad_norm": 0.8671753406524658,
+      "learning_rate": 9.512053701784329e-05,
+      "loss": 1.5776,
+      "step": 410
+    },
+    {
+      "epoch": 0.23131052741587124,
+      "grad_norm": 0.8599863052368164,
+      "learning_rate": 9.490886439642081e-05,
+      "loss": 1.5758,
+      "step": 415
+    },
+    {
+      "epoch": 0.23409740124015885,
+      "grad_norm": 0.9010843634605408,
+      "learning_rate": 9.469294339505098e-05,
+      "loss": 1.6157,
+      "step": 420
+    },
+    {
+      "epoch": 0.23688427506444645,
+      "grad_norm": 0.8823115229606628,
+      "learning_rate": 9.447279443986716e-05,
+      "loss": 1.5334,
+      "step": 425
+    },
+    {
+      "epoch": 0.23967114888873406,
+      "grad_norm": 0.9563323855400085,
+      "learning_rate": 9.424843835696724e-05,
+      "loss": 1.5605,
+      "step": 430
+    },
+    {
+      "epoch": 0.24245802271302166,
+      "grad_norm": 0.9097768068313599,
+      "learning_rate": 9.401989637044355e-05,
+      "loss": 1.5316,
+      "step": 435
+    },
+    {
+      "epoch": 0.24524489653730927,
+      "grad_norm": 0.8611193299293518,
+      "learning_rate": 9.3787190100375e-05,
+      "loss": 1.5611,
+      "step": 440
+    },
+    {
+      "epoch": 0.24803177036159688,
+      "grad_norm": 0.8489307761192322,
+      "learning_rate": 9.355034156078188e-05,
+      "loss": 1.5352,
+      "step": 445
+    },
+    {
+      "epoch": 0.2508186441858845,
+      "grad_norm": 0.8662014603614807,
+      "learning_rate": 9.330937315754329e-05,
+      "loss": 1.5459,
+      "step": 450
+    },
+    {
+      "epoch": 0.2536055180101721,
+      "grad_norm": 0.8968704342842102,
+      "learning_rate": 9.306430768627753e-05,
+      "loss": 1.5309,
+      "step": 455
+    },
+    {
+      "epoch": 0.2563923918344597,
+      "grad_norm": 0.915246844291687,
+      "learning_rate": 9.281516833018571e-05,
+      "loss": 1.552,
+      "step": 460
+    },
+    {
+      "epoch": 0.2591792656587473,
+      "grad_norm": 0.913821280002594,
+      "learning_rate": 9.256197865785854e-05,
+      "loss": 1.6284,
+      "step": 465
+    },
+    {
+      "epoch": 0.26196613948303493,
+      "grad_norm": 0.8203092217445374,
+      "learning_rate": 9.230476262104677e-05,
+      "loss": 1.5235,
+      "step": 470
+    },
+    {
+      "epoch": 0.26475301330732254,
+      "grad_norm": 0.8390527963638306,
+      "learning_rate": 9.204354455239539e-05,
+      "loss": 1.5721,
+      "step": 475
+    },
+    {
+      "epoch": 0.26753988713161014,
+      "grad_norm": 0.8474882245063782,
+      "learning_rate": 9.177834916314165e-05,
+      "loss": 1.5674,
+      "step": 480
+    },
+    {
+      "epoch": 0.27032676095589775,
+      "grad_norm": 0.8627411127090454,
+      "learning_rate": 9.150920154077754e-05,
+      "loss": 1.5457,
+      "step": 485
+    },
+    {
+      "epoch": 0.27311363478018535,
+      "grad_norm": 0.9049586057662964,
+      "learning_rate": 9.123612714667634e-05,
+      "loss": 1.5827,
+      "step": 490
+    },
+    {
+      "epoch": 0.2759005086044729,
+      "grad_norm": 0.806458055973053,
+      "learning_rate": 9.095915181368412e-05,
+      "loss": 1.6398,
+      "step": 495
+    },
+    {
+      "epoch": 0.2786873824287605,
+      "grad_norm": 0.9326282143592834,
+      "learning_rate": 9.067830174367586e-05,
+      "loss": 1.6183,
+      "step": 500
+    },
+    {
+      "epoch": 0.2814742562530481,
+      "grad_norm": 0.8715500831604004,
+      "learning_rate": 9.039360350507679e-05,
+      "loss": 1.5529,
+      "step": 505
+    },
+    {
+      "epoch": 0.2842611300773357,
+      "grad_norm": 0.9053698778152466,
+      "learning_rate": 9.010508403034898e-05,
+      "loss": 1.5207,
+      "step": 510
+    },
+    {
+      "epoch": 0.2870480039016233,
+      "grad_norm": 0.8616856932640076,
+      "learning_rate": 8.98127706134436e-05,
+      "loss": 1.5696,
+      "step": 515
+    },
+    {
+      "epoch": 0.28983487772591093,
+      "grad_norm": 0.9130306243896484,
+      "learning_rate": 8.951669090721881e-05,
+      "loss": 1.5702,
+      "step": 520
+    },
+    {
+      "epoch": 0.29262175155019854,
+      "grad_norm": 0.8649139404296875,
+      "learning_rate": 8.921687292082393e-05,
+      "loss": 1.4826,
+      "step": 525
+    },
+    {
+      "epoch": 0.29540862537448614,
+      "grad_norm": 0.8023464679718018,
+      "learning_rate": 8.891334501704962e-05,
+      "loss": 1.5544,
+      "step": 530
+    },
+    {
+      "epoch": 0.29819549919877375,
+      "grad_norm": 0.9166476726531982,
+      "learning_rate": 8.86061359096449e-05,
+      "loss": 1.5016,
+      "step": 535
+    },
+    {
+      "epoch": 0.30098237302306136,
+      "grad_norm": 0.8918821811676025,
+      "learning_rate": 8.829527466060072e-05,
+      "loss": 1.416,
+      "step": 540
+    },
+    {
+      "epoch": 0.30376924684734896,
+      "grad_norm": 0.8663053512573242,
+      "learning_rate": 8.798079067740077e-05,
+      "loss": 1.641,
+      "step": 545
+    },
+    {
+      "epoch": 0.30655612067163657,
+      "grad_norm": 0.9275104403495789,
+      "learning_rate": 8.766271371023948e-05,
+      "loss": 1.5318,
+      "step": 550
+    },
+    {
+      "epoch": 0.3093429944959242,
+      "grad_norm": 0.9378035664558411,
+      "learning_rate": 8.73410738492077e-05,
+      "loss": 1.4886,
+      "step": 555
+    },
+    {
+      "epoch": 0.3121298683202118,
+      "grad_norm": 0.8419756889343262,
+      "learning_rate": 8.701590152144612e-05,
+      "loss": 1.5012,
+      "step": 560
+    },
+    {
+      "epoch": 0.3149167421444994,
+      "grad_norm": 0.8785234093666077,
+      "learning_rate": 8.668722748826693e-05,
+      "loss": 1.5278,
+      "step": 565
+    },
+    {
+      "epoch": 0.317703615968787,
+      "grad_norm": 0.9210188984870911,
+      "learning_rate": 8.635508284224371e-05,
+      "loss": 1.6802,
+      "step": 570
+    },
+    {
+      "epoch": 0.3204904897930746,
+      "grad_norm": 0.8633784651756287,
+      "learning_rate": 8.601949900427016e-05,
+      "loss": 1.5038,
+      "step": 575
+    },
+    {
+      "epoch": 0.3232773636173622,
+      "grad_norm": 0.9098255634307861,
+      "learning_rate": 8.568050772058762e-05,
+      "loss": 1.5611,
+      "step": 580
+    },
+    {
+      "epoch": 0.3260642374416498,
+      "grad_norm": 0.9021430015563965,
+      "learning_rate": 8.533814105978191e-05,
+      "loss": 1.5166,
+      "step": 585
+    },
+    {
+      "epoch": 0.3288511112659374,
+      "grad_norm": 0.8397455215454102,
+      "learning_rate": 8.499243140974966e-05,
+      "loss": 1.5493,
+      "step": 590
+    },
+    {
+      "epoch": 0.331637985090225,
+      "grad_norm": 0.8716102838516235,
+      "learning_rate": 8.464341147463431e-05,
+      "loss": 1.5242,
+      "step": 595
+    },
+    {
+      "epoch": 0.3344248589145126,
+      "grad_norm": 0.8432428240776062,
+      "learning_rate": 8.429111427173241e-05,
+      "loss": 1.4659,
+      "step": 600
+    },
+    {
+      "epoch": 0.33721173273880023,
+      "grad_norm": 0.8041474223136902,
+      "learning_rate": 8.393557312837018e-05,
+      "loss": 1.4995,
+      "step": 605
+    },
+    {
+      "epoch": 0.33999860656308784,
+      "grad_norm": 0.8596203327178955,
+      "learning_rate": 8.357682167875062e-05,
+      "loss": 1.5639,
+      "step": 610
+    },
+    {
+      "epoch": 0.34278548038737544,
+      "grad_norm": 0.888679027557373,
+      "learning_rate": 8.321489386077192e-05,
+      "loss": 1.5269,
+      "step": 615
+    },
+    {
+      "epoch": 0.34557235421166305,
+      "grad_norm": 0.9395999312400818,
+      "learning_rate": 8.28498239128167e-05,
+      "loss": 1.5474,
+      "step": 620
+    },
+    {
+      "epoch": 0.34835922803595065,
+      "grad_norm": 0.8320375084877014,
+      "learning_rate": 8.248164637051321e-05,
+      "loss": 1.4813,
+      "step": 625
+    },
+    {
+      "epoch": 0.35114610186023826,
+      "grad_norm": 0.901525616645813,
+      "learning_rate": 8.211039606346826e-05,
+      "loss": 1.5674,
+      "step": 630
+    },
+    {
+      "epoch": 0.35393297568452586,
+      "grad_norm": 0.8870389461517334,
+      "learning_rate": 8.173610811197226e-05,
+      "loss": 1.5355,
+      "step": 635
+    },
+    {
+      "epoch": 0.35671984950881347,
+      "grad_norm": 0.8773472309112549,
+      "learning_rate": 8.135881792367686e-05,
+      "loss": 1.5594,
+      "step": 640
+    },
+    {
+      "epoch": 0.3595067233331011,
+      "grad_norm": 0.8549400568008423,
+      "learning_rate": 8.097856119024545e-05,
+      "loss": 1.6111,
+      "step": 645
+    },
+    {
+      "epoch": 0.3622935971573887,
+      "grad_norm": 0.8216549754142761,
+      "learning_rate": 8.059537388397665e-05,
+      "loss": 1.5412,
+      "step": 650
+    },
+    {
+      "epoch": 0.3650804709816763,
+      "grad_norm": 0.8757383227348328,
+      "learning_rate": 8.020929225440137e-05,
+      "loss": 1.544,
+      "step": 655
+    },
+    {
+      "epoch": 0.3678673448059639,
+      "grad_norm": 0.8638432025909424,
+      "learning_rate": 7.98203528248536e-05,
+      "loss": 1.5394,
+      "step": 660
+    },
+    {
+      "epoch": 0.3706542186302515,
+      "grad_norm": 0.9091057181358337,
+      "learning_rate": 7.942859238901528e-05,
+      "loss": 1.5911,
+      "step": 665
+    },
+    {
+      "epoch": 0.3734410924545391,
+      "grad_norm": 1.0159931182861328,
+      "learning_rate": 7.903404800743564e-05,
+      "loss": 1.5694,
+      "step": 670
+    },
+    {
+      "epoch": 0.3762279662788267,
+      "grad_norm": 0.9712515473365784,
+      "learning_rate": 7.863675700402526e-05,
+      "loss": 1.5181,
+      "step": 675
+    },
+    {
+      "epoch": 0.3790148401031143,
+      "grad_norm": 0.9262729287147522,
+      "learning_rate": 7.823675696252524e-05,
+      "loss": 1.604,
+      "step": 680
+    },
+    {
+      "epoch": 0.3818017139274019,
+      "grad_norm": 0.8269544839859009,
+      "learning_rate": 7.783408572295174e-05,
+      "loss": 1.5848,
+      "step": 685
+    },
+    {
+      "epoch": 0.3845885877516895,
+      "grad_norm": 0.8753747344017029,
+      "learning_rate": 7.742878137801639e-05,
+      "loss": 1.5112,
+      "step": 690
+    },
+    {
+      "epoch": 0.38737546157597713,
+      "grad_norm": 0.834754228591919,
+      "learning_rate": 7.702088226952258e-05,
+      "loss": 1.4967,
+      "step": 695
+    },
+    {
+      "epoch": 0.39016233540026474,
+      "grad_norm": 0.9299247860908508,
+      "learning_rate": 7.661042698473853e-05,
+      "loss": 1.4628,
+      "step": 700
+    },
+    {
+      "epoch": 0.39294920922455234,
+      "grad_norm": 0.8204703330993652,
+      "learning_rate": 7.619745435274667e-05,
+      "loss": 1.493,
+      "step": 705
+    },
+    {
+      "epoch": 0.39573608304883995,
+      "grad_norm": 0.9027642011642456,
+      "learning_rate": 7.578200344077073e-05,
+      "loss": 1.5512,
+      "step": 710
+    },
+    {
+      "epoch": 0.39852295687312755,
+      "grad_norm": 0.923340380191803,
+      "learning_rate": 7.536411355047964e-05,
+      "loss": 1.5396,
+      "step": 715
+    },
+    {
+      "epoch": 0.40130983069741516,
+      "grad_norm": 0.855146586894989,
+      "learning_rate": 7.494382421426984e-05,
+      "loss": 1.5894,
+      "step": 720
+    },
+    {
+      "epoch": 0.40409670452170277,
+      "grad_norm": 0.850355327129364,
+      "learning_rate": 7.452117519152542e-05,
+      "loss": 1.474,
+      "step": 725
+    },
+    {
+      "epoch": 0.40688357834599037,
+      "grad_norm": 0.9606896638870239,
+      "learning_rate": 7.409620646485685e-05,
+      "loss": 1.506,
+      "step": 730
+    },
+    {
+      "epoch": 0.409670452170278,
+      "grad_norm": 0.8371616005897522,
+      "learning_rate": 7.36689582363187e-05,
+      "loss": 1.5454,
+      "step": 735
+    },
+    {
+      "epoch": 0.4124573259945656,
+      "grad_norm": 0.8478779792785645,
+      "learning_rate": 7.323947092360649e-05,
+      "loss": 1.4607,
+      "step": 740
+    },
+    {
+      "epoch": 0.4152441998188532,
+      "grad_norm": 0.9579894542694092,
+      "learning_rate": 7.280778515623314e-05,
+      "loss": 1.5948,
+      "step": 745
+    },
+    {
+      "epoch": 0.4180310736431408,
+      "grad_norm": 0.9187178015708923,
+      "learning_rate": 7.237394177168548e-05,
+      "loss": 1.5587,
+      "step": 750
+    },
+    {
+      "epoch": 0.4208179474674284,
+      "grad_norm": 0.9753735661506653,
+      "learning_rate": 7.193798181156095e-05,
+      "loss": 1.5907,
+      "step": 755
+    },
+    {
+      "epoch": 0.423604821291716,
+      "grad_norm": 0.8725233674049377,
+      "learning_rate": 7.149994651768514e-05,
+      "loss": 1.4151,
+      "step": 760
+    },
+    {
+      "epoch": 0.4263916951160036,
+      "grad_norm": 0.9570207595825195,
+      "learning_rate": 7.10598773282103e-05,
+      "loss": 1.5483,
+      "step": 765
+    },
+    {
+      "epoch": 0.4291785689402912,
+      "grad_norm": 0.9618322253227234,
+      "learning_rate": 7.061781587369519e-05,
+      "loss": 1.6014,
+      "step": 770
+    },
+    {
+      "epoch": 0.4319654427645788,
+      "grad_norm": 0.8392998576164246,
+      "learning_rate": 7.017380397316695e-05,
+      "loss": 1.4826,
+      "step": 775
+    },
+    {
+      "epoch": 0.4347523165888664,
+      "grad_norm": 0.8500150442123413,
+      "learning_rate": 6.972788363016497e-05,
+      "loss": 1.4199,
+      "step": 780
+    },
+    {
+      "epoch": 0.43753919041315403,
+      "grad_norm": 0.9263842105865479,
+      "learning_rate": 6.92800970287674e-05,
+      "loss": 1.5328,
+      "step": 785
+    },
+    {
+      "epoch": 0.44032606423744164,
+      "grad_norm": 0.8920561671257019,
+      "learning_rate": 6.883048652960038e-05,
+      "loss": 1.5331,
+      "step": 790
+    },
+    {
+      "epoch": 0.44311293806172924,
+      "grad_norm": 0.8388264775276184,
+      "learning_rate": 6.837909466583095e-05,
+      "loss": 1.5442,
+      "step": 795
+    },
+    {
+      "epoch": 0.44589981188601685,
+      "grad_norm": 0.8242114186286926,
+      "learning_rate": 6.792596413914324e-05,
+      "loss": 1.4163,
+      "step": 800
+    },
+    {
+      "epoch": 0.44868668571030446,
+      "grad_norm": 0.9328235387802124,
+      "learning_rate": 6.747113781569892e-05,
+      "loss": 1.5607,
+      "step": 805
+    },
+    {
+      "epoch": 0.45147355953459206,
+      "grad_norm": 0.9171831607818604,
+      "learning_rate": 6.701465872208216e-05,
+      "loss": 1.5812,
+      "step": 810
+    },
+    {
+      "epoch": 0.45426043335887967,
+      "grad_norm": 0.9111394882202148,
+      "learning_rate": 6.655657004122916e-05,
+      "loss": 1.5764,
+      "step": 815
+    },
+    {
+      "epoch": 0.4570473071831673,
+      "grad_norm": 0.9669636487960815,
+      "learning_rate": 6.60969151083432e-05,
+      "loss": 1.5125,
+      "step": 820
+    },
+    {
+      "epoch": 0.4598341810074549,
+      "grad_norm": 0.8893470168113708,
+      "learning_rate": 6.563573740679496e-05,
+      "loss": 1.5136,
+      "step": 825
+    },
+    {
+      "epoch": 0.4626210548317425,
+      "grad_norm": 0.8966957926750183,
+      "learning_rate": 6.517308056400917e-05,
+      "loss": 1.5365,
+      "step": 830
+    },
+    {
+      "epoch": 0.4654079286560301,
+      "grad_norm": 0.9211333990097046,
+      "learning_rate": 6.470898834733731e-05,
+      "loss": 1.5054,
+      "step": 835
+    },
+    {
+      "epoch": 0.4681948024803177,
+      "grad_norm": 0.8702636361122131,
+      "learning_rate": 6.42435046599173e-05,
+      "loss": 1.5078,
+      "step": 840
+    },
+    {
+      "epoch": 0.4709816763046053,
+      "grad_norm": 1.0185481309890747,
+      "learning_rate": 6.377667353652022e-05,
+      "loss": 1.5513,
+      "step": 845
+    },
+    {
+      "epoch": 0.4737685501288929,
+      "grad_norm": 0.8714690208435059,
+      "learning_rate": 6.330853913938466e-05,
+      "loss": 1.4742,
+      "step": 850
+    },
+    {
+      "epoch": 0.4765554239531805,
+      "grad_norm": 0.8962568044662476,
+      "learning_rate": 6.283914575403888e-05,
+      "loss": 1.5775,
+      "step": 855
+    },
+    {
+      "epoch": 0.4793422977774681,
+      "grad_norm": 0.9320731163024902,
+      "learning_rate": 6.236853778511156e-05,
+      "loss": 1.5052,
+      "step": 860
+    },
+    {
+      "epoch": 0.4821291716017557,
+      "grad_norm": 0.9507172703742981,
+      "learning_rate": 6.189675975213094e-05,
+      "loss": 1.5255,
+      "step": 865
+    },
+    {
+      "epoch": 0.48491604542604333,
+      "grad_norm": 0.9428796172142029,
+      "learning_rate": 6.142385628531342e-05,
+      "loss": 1.5391,
+      "step": 870
+    },
+    {
+      "epoch": 0.48770291925033094,
+      "grad_norm": 0.9637115001678467,
+      "learning_rate": 6.09498721213414e-05,
+      "loss": 1.4856,
+      "step": 875
+    },
+    {
+      "epoch": 0.49048979307461854,
+      "grad_norm": 0.9000864624977112,
+      "learning_rate": 6.047485209913137e-05,
+      "loss": 1.6011,
+      "step": 880
+    },
+    {
+      "epoch": 0.49327666689890615,
+      "grad_norm": 0.820088803768158,
+      "learning_rate": 5.999884115559192e-05,
+      "loss": 1.4913,
+      "step": 885
+    },
+    {
+      "epoch": 0.49606354072319375,
+      "grad_norm": 0.7902366518974304,
+      "learning_rate": 5.952188432137293e-05,
+      "loss": 1.5304,
+      "step": 890
+    },
+    {
+      "epoch": 0.49885041454748136,
+      "grad_norm": 0.7885243892669678,
+      "learning_rate": 5.90440267166055e-05,
+      "loss": 1.4527,
+      "step": 895
+    },
+    {
+      "epoch": 0.501637288371769,
+      "grad_norm": 1.0124362707138062,
+      "learning_rate": 5.8565313546633684e-05,
+      "loss": 1.5122,
+      "step": 900
+    },
+    {
+      "epoch": 0.5044241621960566,
+      "grad_norm": 0.8640070557594299,
+      "learning_rate": 5.8085790097738025e-05,
+      "loss": 1.5547,
+      "step": 905
+    },
+    {
+      "epoch": 0.5072110360203442,
+      "grad_norm": 0.9227625131607056,
+      "learning_rate": 5.7605501732851475e-05,
+      "loss": 1.5804,
+      "step": 910
+    },
+    {
+      "epoch": 0.5099979098446318,
+      "grad_norm": 0.8405686616897583,
+      "learning_rate": 5.712449388726807e-05,
+      "loss": 1.5661,
+      "step": 915
+    },
+    {
+      "epoch": 0.5127847836689194,
+      "grad_norm": 0.867641031742096,
+      "learning_rate": 5.664281206434472e-05,
+      "loss": 1.4674,
+      "step": 920
+    },
+    {
+      "epoch": 0.515571657493207,
+      "grad_norm": 0.8705233931541443,
+      "learning_rate": 5.616050183119663e-05,
+      "loss": 1.5702,
+      "step": 925
+    },
+    {
+      "epoch": 0.5183585313174947,
+      "grad_norm": 0.8183286786079407,
+      "learning_rate": 5.5677608814386616e-05,
+      "loss": 1.3906,
+      "step": 930
+    },
+    {
+      "epoch": 0.5211454051417822,
+      "grad_norm": 0.9858644604682922,
+      "learning_rate": 5.519417869560889e-05,
+      "loss": 1.5129,
+      "step": 935
+    },
+    {
+      "epoch": 0.5239322789660699,
+      "grad_norm": 0.9389538168907166,
+      "learning_rate": 5.471025720736747e-05,
+      "loss": 1.55,
+      "step": 940
+    },
+    {
+      "epoch": 0.5267191527903574,
+      "grad_norm": 0.8737983107566833,
+      "learning_rate": 5.422589012864996e-05,
+      "loss": 1.5582,
+      "step": 945
+    },
+    {
+      "epoch": 0.5295060266146451,
+      "grad_norm": 0.9484609961509705,
+      "learning_rate": 5.3741123280596864e-05,
+      "loss": 1.4895,
+      "step": 950
+    },
+    {
+      "epoch": 0.5322929004389326,
+      "grad_norm": 0.9373424649238586,
+      "learning_rate": 5.325600252216685e-05,
+      "loss": 1.502,
+      "step": 955
+    },
+    {
+      "epoch": 0.5350797742632203,
+      "grad_norm": 0.9036156535148621,
+      "learning_rate": 5.27705737457985e-05,
+      "loss": 1.394,
+      "step": 960
+    },
+    {
+      "epoch": 0.5378666480875078,
+      "grad_norm": 0.9241552948951721,
+      "learning_rate": 5.228488287306896e-05,
+      "loss": 1.5101,
+      "step": 965
+    },
+    {
+      "epoch": 0.5406535219117955,
+      "grad_norm": 0.8161901235580444,
+      "learning_rate": 5.179897585034963e-05,
+      "loss": 1.5254,
+      "step": 970
+    },
+    {
+      "epoch": 0.543440395736083,
+      "grad_norm": 0.8650552034378052,
+      "learning_rate": 5.1312898644459776e-05,
+      "loss": 1.4365,
+      "step": 975
+    },
+    {
+      "epoch": 0.5462272695603707,
+      "grad_norm": 0.9877922534942627,
+      "learning_rate": 5.0826697238317935e-05,
+      "loss": 1.5132,
+      "step": 980
+    },
+    {
+      "epoch": 0.5490141433846583,
+      "grad_norm": 0.8969970941543579,
+      "learning_rate": 5.0340417626592016e-05,
+      "loss": 1.551,
+      "step": 985
+    },
+    {
+      "epoch": 0.5518010172089458,
+      "grad_norm": 0.9135898351669312,
+      "learning_rate": 4.9854105811348216e-05,
+      "loss": 1.5476,
+      "step": 990
+    },
+    {
+      "epoch": 0.5545878910332335,
+      "grad_norm": 0.9921842217445374,
+      "learning_rate": 4.936780779769913e-05,
+      "loss": 1.5394,
+      "step": 995
+    },
+    {
+      "epoch": 0.557374764857521,
+      "grad_norm": 0.8833547830581665,
+      "learning_rate": 4.888156958945174e-05,
+      "loss": 1.5002,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5601616386818087,
+      "grad_norm": 0.9588260650634766,
+      "learning_rate": 4.839543718475543e-05,
+      "loss": 1.4957,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5629485125060962,
+      "grad_norm": 0.9499724507331848,
+      "learning_rate": 4.790945657175061e-05,
+      "loss": 1.4931,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5657353863303839,
+      "grad_norm": 0.9982417821884155,
+      "learning_rate": 4.742367372421811e-05,
+      "loss": 1.5432,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5685222601546714,
+      "grad_norm": 0.9785814881324768,
+      "learning_rate": 4.69381345972302e-05,
+      "loss": 1.4732,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5713091339789591,
+      "grad_norm": 0.8655848503112793,
+      "learning_rate": 4.6452885122803205e-05,
+      "loss": 1.4509,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5740960078032467,
+      "grad_norm": 0.8517627120018005,
+      "learning_rate": 4.5967971205552194e-05,
+      "loss": 1.5168,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5768828816275343,
+      "grad_norm": 0.8920270800590515,
+      "learning_rate": 4.548343871834864e-05,
+      "loss": 1.5045,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5796697554518219,
+      "grad_norm": 0.8877922892570496,
+      "learning_rate": 4.499933349798067e-05,
+      "loss": 1.5649,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5824566292761095,
+      "grad_norm": 0.9257612824440002,
+      "learning_rate": 4.451570134081694e-05,
+      "loss": 1.6024,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5852435031003971,
+      "grad_norm": 0.8532921075820923,
+      "learning_rate": 4.403258799847433e-05,
+      "loss": 1.4751,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5880303769246847,
+      "grad_norm": 0.9549328088760376,
+      "learning_rate": 4.3550039173489845e-05,
+      "loss": 1.5097,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5908172507489723,
+      "grad_norm": 0.935123860836029,
+      "learning_rate": 4.306810051499708e-05,
+      "loss": 1.4394,
+      "step": 1060
+    },
+    {
+      "epoch": 0.59360412457326,
+      "grad_norm": 0.9845610857009888,
+      "learning_rate": 4.2586817614407895e-05,
+      "loss": 1.5116,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5963909983975475,
+      "grad_norm": 0.7877358198165894,
+      "learning_rate": 4.210623600109946e-05,
+      "loss": 1.3753,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5991778722218352,
+      "grad_norm": 0.8905171155929565,
+      "learning_rate": 4.162640113810706e-05,
+      "loss": 1.5319,
+      "step": 1075
+    },
+    {
+      "epoch": 0.6019647460461227,
+      "grad_norm": 0.9214166402816772,
+      "learning_rate": 4.114735841782347e-05,
+      "loss": 1.5835,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6047516198704104,
+      "grad_norm": 0.9513778686523438,
+      "learning_rate": 4.06691531577047e-05,
+      "loss": 1.5054,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6075384936946979,
+      "grad_norm": 0.8654255270957947,
+      "learning_rate": 4.019183059598296e-05,
+      "loss": 1.4682,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6103253675189856,
+      "grad_norm": 0.8639852404594421,
+      "learning_rate": 3.971543588738724e-05,
+      "loss": 1.4683,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6131122413432731,
+      "grad_norm": 0.9456416368484497,
+      "learning_rate": 3.924001409887158e-05,
+      "loss": 1.4401,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6158991151675608,
+      "grad_norm": 0.9741712212562561,
+      "learning_rate": 3.87656102053517e-05,
+      "loss": 1.5094,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6186859889918483,
+      "grad_norm": 0.8833429217338562,
+      "learning_rate": 3.8292269085450474e-05,
+      "loss": 1.4824,
+      "step": 1110
+    },
+    {
+      "epoch": 0.621472862816136,
+      "grad_norm": 0.8431907892227173,
+      "learning_rate": 3.782003551725236e-05,
+      "loss": 1.5589,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6242597366404236,
+      "grad_norm": 0.8895665407180786,
+      "learning_rate": 3.734895417406734e-05,
+      "loss": 1.5259,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6270466104647112,
+      "grad_norm": 0.9013938903808594,
+      "learning_rate": 3.687906962020491e-05,
+      "loss": 1.5298,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6298334842889988,
+      "grad_norm": 0.9939606189727783,
+      "learning_rate": 3.641042630675829e-05,
+      "loss": 1.5316,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6326203581132864,
+      "grad_norm": 0.8754637837409973,
+      "learning_rate": 3.594306856739924e-05,
+      "loss": 1.4538,
+      "step": 1135
+    },
+    {
+      "epoch": 0.635407231937574,
+      "grad_norm": 0.8571580648422241,
+      "learning_rate": 3.547704061418424e-05,
+      "loss": 1.4824,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6381941057618616,
+      "grad_norm": 0.9635239839553833,
+      "learning_rate": 3.501238653337194e-05,
+      "loss": 1.4903,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6409809795861492,
+      "grad_norm": 0.9419766068458557,
+      "learning_rate": 3.4549150281252636e-05,
+      "loss": 1.5533,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6437678534104369,
+      "grad_norm": 0.8649463653564453,
+      "learning_rate": 3.408737567998993e-05,
+      "loss": 1.3963,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6465547272347244,
+      "grad_norm": 0.9010219573974609,
+      "learning_rate": 3.362710641347524e-05,
+      "loss": 1.5465,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6493416010590121,
+      "grad_norm": 0.8567179441452026,
+      "learning_rate": 3.316838602319532e-05,
+      "loss": 1.5415,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6521284748832996,
+      "grad_norm": 0.9118956327438354,
+      "learning_rate": 3.271125790411309e-05,
+      "loss": 1.3886,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6549153487075873,
+      "grad_norm": 1.0297619104385376,
+      "learning_rate": 3.225576530056264e-05,
+      "loss": 1.5354,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6577022225318748,
+      "grad_norm": 0.8590645790100098,
+      "learning_rate": 3.180195130215824e-05,
+      "loss": 1.4846,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6604890963561625,
+      "grad_norm": 1.0237464904785156,
+      "learning_rate": 3.1349858839717986e-05,
+      "loss": 1.4685,
+      "step": 1185
+    },
+    {
+      "epoch": 0.66327597018045,
+      "grad_norm": 0.9131646752357483,
+      "learning_rate": 3.089953068120271e-05,
+      "loss": 1.4137,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6660628440047377,
+      "grad_norm": 0.9120528101921082,
+      "learning_rate": 3.0451009427669986e-05,
+      "loss": 1.4465,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6688497178290252,
+      "grad_norm": 0.9107281565666199,
+      "learning_rate": 3.000433750924414e-05,
+      "loss": 1.4854,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6716365916533129,
+      "grad_norm": 0.870647668838501,
+      "learning_rate": 2.9559557181102315e-05,
+      "loss": 1.4992,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6744234654776005,
+      "grad_norm": 0.9665882587432861,
+      "learning_rate": 2.911671051947722e-05,
+      "loss": 1.4342,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6772103393018881,
+      "grad_norm": 0.9868600368499756,
+      "learning_rate": 2.867583941767657e-05,
+      "loss": 1.4412,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6799972131261757,
+      "grad_norm": 0.8998962640762329,
+      "learning_rate": 2.823698558212009e-05,
+      "loss": 1.497,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6827840869504633,
+      "grad_norm": 0.9534763693809509,
+      "learning_rate": 2.7800190528394122e-05,
+      "loss": 1.4519,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6855709607747509,
+      "grad_norm": 0.8596827983856201,
+      "learning_rate": 2.736549557732405e-05,
+      "loss": 1.4384,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6883578345990385,
+      "grad_norm": 0.9013137817382812,
+      "learning_rate": 2.693294185106562e-05,
+      "loss": 1.6527,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6911447084233261,
+      "grad_norm": 0.9170345067977905,
+      "learning_rate": 2.650257026921455e-05,
+      "loss": 1.5202,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6939315822476138,
+      "grad_norm": 0.9849944710731506,
+      "learning_rate": 2.607442154493568e-05,
+      "loss": 1.5388,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6967184560719013,
+      "grad_norm": 0.9363320469856262,
+      "learning_rate": 2.5648536181111438e-05,
+      "loss": 1.488,
+      "step": 1250
+    },
+    {
+      "epoch": 0.699505329896189,
+      "grad_norm": 0.9596914052963257,
+      "learning_rate": 2.5224954466510274e-05,
+      "loss": 1.4646,
+      "step": 1255
+    },
+    {
+      "epoch": 0.7022922037204765,
+      "grad_norm": 0.9537048935890198,
+      "learning_rate": 2.480371647197538e-05,
+      "loss": 1.5223,
+      "step": 1260
+    },
+    {
+      "epoch": 0.7050790775447642,
+      "grad_norm": 0.9345447421073914,
+      "learning_rate": 2.438486204663391e-05,
+      "loss": 1.5393,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7078659513690517,
+      "grad_norm": 0.9050605893135071,
+      "learning_rate": 2.3968430814127385e-05,
+      "loss": 1.5276,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7106528251933394,
+      "grad_norm": 0.9313639402389526,
+      "learning_rate": 2.3554462168863085e-05,
+      "loss": 1.507,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7134396990176269,
+      "grad_norm": 0.9372873902320862,
+      "learning_rate": 2.314299527228759e-05,
+      "loss": 1.5198,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7162265728419146,
+      "grad_norm": 0.8948291540145874,
+      "learning_rate": 2.2734069049181882e-05,
+      "loss": 1.4833,
+      "step": 1285
+    },
+    {
+      "epoch": 0.7190134466662021,
+      "grad_norm": 0.9705023169517517,
+      "learning_rate": 2.2327722183979212e-05,
+      "loss": 1.5114,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7218003204904898,
+      "grad_norm": 0.9261203408241272,
+      "learning_rate": 2.1923993117105462e-05,
+      "loss": 1.4665,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7245871943147774,
+      "grad_norm": 1.0300346612930298,
+      "learning_rate": 2.1522920041342704e-05,
+      "loss": 1.4931,
+      "step": 1300
+    },
+    {
+      "epoch": 0.727374068139065,
+      "grad_norm": 0.970912754535675,
+      "learning_rate": 2.1124540898216248e-05,
+      "loss": 1.5079,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7301609419633526,
+      "grad_norm": 0.9137006402015686,
+      "learning_rate": 2.0728893374405166e-05,
+      "loss": 1.4083,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7329478157876402,
+      "grad_norm": 1.0079381465911865,
+      "learning_rate": 2.033601489817738e-05,
+      "loss": 1.6,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7357346896119278,
+      "grad_norm": 1.039553165435791,
+      "learning_rate": 1.9945942635848748e-05,
+      "loss": 1.5251,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7385215634362154,
+      "grad_norm": 0.9428099393844604,
+      "learning_rate": 1.9558713488267238e-05,
+      "loss": 1.5303,
+      "step": 1325
+    },
+    {
+      "epoch": 0.741308437260503,
+      "grad_norm": 0.9896711707115173,
+      "learning_rate": 1.917436408732208e-05,
+      "loss": 1.4847,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7440953110847907,
+      "grad_norm": 1.0423614978790283,
+      "learning_rate": 1.8792930792478357e-05,
+      "loss": 1.4854,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7468821849090782,
+      "grad_norm": 1.0262562036514282,
+      "learning_rate": 1.8414449687337464e-05,
+      "loss": 1.4347,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7496690587333659,
+      "grad_norm": 0.9324917793273926,
+      "learning_rate": 1.8038956576223504e-05,
+      "loss": 1.4126,
+      "step": 1345
+    },
+    {
+      "epoch": 0.7524559325576534,
+      "grad_norm": 0.8801437616348267,
+      "learning_rate": 1.766648698079635e-05,
+      "loss": 1.5037,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7552428063819411,
+      "grad_norm": 0.9760573506355286,
+      "learning_rate": 1.7297076136691072e-05,
+      "loss": 1.5874,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7580296802062286,
+      "grad_norm": 1.0139156579971313,
+      "learning_rate": 1.6930758990184875e-05,
+      "loss": 1.5052,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7608165540305163,
+      "grad_norm": 1.0769689083099365,
+      "learning_rate": 1.6567570194891024e-05,
+      "loss": 1.5064,
+      "step": 1365
+    },
+    {
+      "epoch": 0.7636034278548038,
+      "grad_norm": 1.0925158262252808,
+      "learning_rate": 1.620754410848069e-05,
+      "loss": 1.497,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7663903016790915,
+      "grad_norm": 0.9047788977622986,
+      "learning_rate": 1.5850714789432663e-05,
+      "loss": 1.5339,
+      "step": 1375
+    },
+    {
+      "epoch": 0.769177175503379,
+      "grad_norm": 1.0132824182510376,
+      "learning_rate": 1.549711599381145e-05,
+      "loss": 1.5058,
+      "step": 1380
+    },
+    {
+      "epoch": 0.7719640493276667,
+      "grad_norm": 0.8855955004692078,
+      "learning_rate": 1.5146781172073959e-05,
+      "loss": 1.465,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7747509231519543,
+      "grad_norm": 0.9173756837844849,
+      "learning_rate": 1.479974346590503e-05,
+      "loss": 1.469,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7775377969762419,
+      "grad_norm": 1.0163615942001343,
+      "learning_rate": 1.4456035705082349e-05,
+      "loss": 1.477,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7803246708005295,
+      "grad_norm": 0.9795538187026978,
+      "learning_rate": 1.4115690404370551e-05,
+      "loss": 1.4778,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7831115446248171,
+      "grad_norm": 0.9054852724075317,
+      "learning_rate": 1.3778739760445552e-05,
+      "loss": 1.4847,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7858984184491047,
+      "grad_norm": 0.9668756127357483,
+      "learning_rate": 1.344521564884858e-05,
+      "loss": 1.549,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7886852922733923,
+      "grad_norm": 0.9046013951301575,
+      "learning_rate": 1.3115149620970795e-05,
+      "loss": 1.5333,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7914721660976799,
+      "grad_norm": 0.9518187642097473,
+      "learning_rate": 1.2788572901068552e-05,
+      "loss": 1.467,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7942590399219676,
+      "grad_norm": 1.0664600133895874,
+      "learning_rate": 1.2465516383309551e-05,
+      "loss": 1.5123,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7970459137462551,
+      "grad_norm": 0.9351513385772705,
+      "learning_rate": 1.2146010628850268e-05,
+      "loss": 1.5253,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7998327875705428,
+      "grad_norm": 0.9090898036956787,
+      "learning_rate": 1.183008586294485e-05,
+      "loss": 1.4444,
+      "step": 1435
+    },
+    {
+      "epoch": 0.8026196613948303,
+      "grad_norm": 0.92516028881073,
+      "learning_rate": 1.151777197208585e-05,
+      "loss": 1.4712,
+      "step": 1440
+    },
+    {
+      "epoch": 0.805406535219118,
+      "grad_norm": 0.9981464743614197,
+      "learning_rate": 1.1209098501176896e-05,
+      "loss": 1.462,
+      "step": 1445
+    },
+    {
+      "epoch": 0.8081934090434055,
+      "grad_norm": 0.9704977869987488,
+      "learning_rate": 1.0904094650737795e-05,
+      "loss": 1.4623,
+      "step": 1450
+    },
+    {
+      "epoch": 0.8109802828676932,
+      "grad_norm": 1.0866062641143799,
+      "learning_rate": 1.0602789274142133e-05,
+      "loss": 1.5243,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8137671566919807,
+      "grad_norm": 0.963045597076416,
+      "learning_rate": 1.0305210874887766e-05,
+      "loss": 1.4094,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8165540305162684,
+      "grad_norm": 0.9393243193626404,
+      "learning_rate": 1.0011387603900385e-05,
+      "loss": 1.4811,
+      "step": 1465
+    },
+    {
+      "epoch": 0.819340904340556,
+      "grad_norm": 0.9286583662033081,
+      "learning_rate": 9.7213472568704e-06,
+      "loss": 1.4956,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8221277781648436,
+      "grad_norm": 0.9984036087989807,
+      "learning_rate": 9.435117271623566e-06,
+      "loss": 1.3546,
+      "step": 1475
+    },
+    {
+      "epoch": 0.8249146519891312,
+      "grad_norm": 0.987972617149353,
+      "learning_rate": 9.152724725525202e-06,
+      "loss": 1.4828,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8277015258134188,
+      "grad_norm": 0.9146081805229187,
+      "learning_rate": 8.87419633291886e-06,
+      "loss": 1.5321,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8304883996377064,
+      "grad_norm": 1.007124662399292,
+      "learning_rate": 8.599558442598998e-06,
+      "loss": 1.4268,
+      "step": 1490
+    },
+    {
+      "epoch": 0.833275273461994,
+      "grad_norm": 1.024166226387024,
+      "learning_rate": 8.328837035318448e-06,
+      "loss": 1.5225,
+      "step": 1495
+    },
+    {
+      "epoch": 0.8360621472862816,
+      "grad_norm": 0.8978021740913391,
+      "learning_rate": 8.06205772133063e-06,
+      "loss": 1.4085,
+      "step": 1500
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 1795,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.525796876090176e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1500/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79ce14c8b9b5fd843768a164624a4f08d081472da21dd3d1eb9a404fc8d77b0d
+size 6161

checkpoint-1795/README.md

@@ -0,0 +1,202 @@
+---
+base_model: google/gemma-3-12b-it
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.15.2

checkpoint-1795/adapter_config.json

@@ -0,0 +1,180 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-3-12b-it",
+  "bias": "none",
+  "corda_config": null,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "language_model.layers.3.self_attn.q_proj",
+    "35.self_attn.k_proj",
+    "o_proj",
+    "40.self_attn.q_proj",
+    "language_model.layers.8.self_attn.k_proj",
+    "language_model.layers.25.self_attn.v_proj",
+    "language_model.layers.11.self_attn.k_proj",
+    "43.self_attn.q_proj",
+    "language_model.layers.15.self_attn.q_proj",
+    "39.self_attn.q_proj",
+    "language_model.layers.21.self_attn.v_proj",
+    "35.self_attn.q_proj",
+    "31.self_attn.v_proj",
+    "language_model.layers.8.self_attn.v_proj",
+    "38.self_attn.k_proj",
+    "language_model.layers.0.self_attn.k_proj",
+    "language_model.layers.12.self_attn.v_proj",
+    "27.self_attn.q_proj",
+    "language_model.layers.5.self_attn.q_proj",
+    "language_model.layers.2.self_attn.q_proj",
+    "language_model.layers.20.self_attn.k_proj",
+    "up_proj",
+    "language_model.layers.14.self_attn.q_proj",
+    "language_model.layers.3.self_attn.v_proj",
+    "language_model.layers.21.self_attn.q_proj",
+    "language_model.layers.26.self_attn.q_proj",
+    "down_proj",
+    "language_model.layers.23.self_attn.q_proj",
+    "language_model.layers.8.self_attn.q_proj",
+    "30.self_attn.v_proj",
+    "language_model.layers.1.self_attn.v_proj",
+    "36.self_attn.q_proj",
+    "28.self_attn.v_proj",
+    "language_model.layers.23.self_attn.v_proj",
+    "language_model.layers.22.self_attn.q_proj",
+    "language_model.layers.1.self_attn.k_proj",
+    "38.self_attn.q_proj",
+    "32.self_attn.q_proj",
+    "46.self_attn.k_proj",
+    "34.self_attn.k_proj",
+    "language_model.layers.16.self_attn.v_proj",
+    "language_model.layers.5.self_attn.v_proj",
+    "47.self_attn.k_proj",
+    "42.self_attn.k_proj",
+    "45.self_attn.k_proj",
+    "45.self_attn.v_proj",
+    "language_model.layers.19.self_attn.k_proj",
+    "language_model.layers.22.self_attn.k_proj",
+    "language_model.layers.17.self_attn.v_proj",
+    "language_model.layers.17.self_attn.q_proj",
+    "language_model.layers.9.self_attn.k_proj",
+    "language_model.layers.6.self_attn.q_proj",
+    "language_model.layers.1.self_attn.q_proj",
+    "language_model.layers.7.self_attn.k_proj",
+    "39.self_attn.k_proj",
+    "language_model.layers.6.self_attn.k_proj",
+    "language_model.layers.18.self_attn.q_proj",
+    "language_model.layers.5.self_attn.k_proj",
+    "29.self_attn.k_proj",
+    "language_model.layers.20.self_attn.v_proj",
+    "language_model.layers.25.self_attn.k_proj",
+    "language_model.layers.4.self_attn.q_proj",
+    "language_model.layers.13.self_attn.k_proj",
+    "41.self_attn.v_proj",
+    "language_model.layers.12.self_attn.q_proj",
+    "language_model.layers.7.self_attn.q_proj",
+    "language_model.layers.19.self_attn.v_proj",
+    "language_model.layers.7.self_attn.v_proj",
+    "language_model.layers.11.self_attn.q_proj",
+    "30.self_attn.k_proj",
+    "language_model.layers.16.self_attn.k_proj",
+    "43.self_attn.k_proj",
+    "47.self_attn.q_proj",
+    "language_model.layers.10.self_attn.v_proj",
+    "language_model.layers.15.self_attn.v_proj",
+    "language_model.layers.0.self_attn.v_proj",
+    "31.self_attn.k_proj",
+    "35.self_attn.v_proj",
+    "language_model.layers.2.self_attn.k_proj",
+    "27.self_attn.v_proj",
+    "30.self_attn.q_proj",
+    "28.self_attn.k_proj",
+    "41.self_attn.k_proj",
+    "language_model.layers.0.self_attn.q_proj",
+    "33.self_attn.v_proj",
+    "language_model.layers.6.self_attn.v_proj",
+    "44.self_attn.q_proj",
+    "language_model.layers.2.self_attn.v_proj",
+    "language_model.layers.3.self_attn.k_proj",
+    "language_model.layers.9.self_attn.q_proj",
+    "37.self_attn.q_proj",
+    "language_model.layers.4.self_attn.v_proj",
+    "33.self_attn.q_proj",
+    "29.self_attn.v_proj",
+    "language_model.layers.18.self_attn.k_proj",
+    "37.self_attn.v_proj",
+    "language_model.layers.17.self_attn.k_proj",
+    "33.self_attn.k_proj",
+    "language_model.layers.19.self_attn.q_proj",
+    "gate_proj",
+    "language_model.layers.10.self_attn.q_proj",
+    "47.self_attn.v_proj",
+    "42.self_attn.q_proj",
+    "language_model.layers.16.self_attn.q_proj",
+    "language_model.layers.25.self_attn.q_proj",
+    "42.self_attn.v_proj",
+    "32.self_attn.k_proj",
+    "language_model.layers.9.self_attn.v_proj",
+    "46.self_attn.v_proj",
+    "language_model.layers.12.self_attn.k_proj",
+    "38.self_attn.v_proj",
+    "language_model.layers.14.self_attn.k_proj",
+    "46.self_attn.q_proj",
+    "language_model.layers.18.self_attn.v_proj",
+    "language_model.layers.26.self_attn.v_proj",
+    "44.self_attn.v_proj",
+    "language_model.layers.24.self_attn.q_proj",
+    "language_model.layers.15.self_attn.k_proj",
+    "language_model.layers.14.self_attn.v_proj",
+    "29.self_attn.q_proj",
+    "language_model.layers.4.self_attn.k_proj",
+    "language_model.layers.13.self_attn.q_proj",
+    "43.self_attn.v_proj",
+    "34.self_attn.q_proj",
+    "45.self_attn.q_proj",
+    "32.self_attn.v_proj",
+    "language_model.layers.11.self_attn.v_proj",
+    "36.self_attn.v_proj",
+    "language_model.layers.21.self_attn.k_proj",
+    "31.self_attn.q_proj",
+    "37.self_attn.k_proj",
+    "39.self_attn.v_proj",
+    "language_model.layers.26.self_attn.k_proj",
+    "40.self_attn.k_proj",
+    "36.self_attn.k_proj",
+    "41.self_attn.q_proj",
+    "language_model.layers.24.self_attn.v_proj",
+    "language_model.layers.23.self_attn.k_proj",
+    "language_model.layers.10.self_attn.k_proj",
+    "27.self_attn.k_proj",
+    "language_model.layers.13.self_attn.v_proj",
+    "34.self_attn.v_proj",
+    "40.self_attn.v_proj",
+    "language_model.layers.20.self_attn.q_proj",
+    "language_model.layers.24.self_attn.k_proj",
+    "44.self_attn.k_proj",
+    "28.self_attn.q_proj",
+    "language_model.layers.22.self_attn.v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-1795/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f605bf90a7e13b22d18038eeb0bffe63d3b68562cec8e726a11e3f9fa51433d
+size 131040264

checkpoint-1795/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<image_soft_token>": 262144
+}

checkpoint-1795/chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

checkpoint-1795/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89444b4b820bb5160048c8d0eb68060f0e64d2e17a8f2755da7edb15445f3f7b
+size 262448707

checkpoint-1795/preprocessor_config.json

@@ -0,0 +1,29 @@
+{
+  "do_convert_rgb": null,
+  "do_normalize": true,
+  "do_pan_and_scan": null,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "Gemma3ImageProcessor",
+  "image_seq_length": 256,
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "pan_and_scan_max_num_crops": null,
+  "pan_and_scan_min_crop_size": null,
+  "pan_and_scan_min_ratio_to_activate": null,
+  "processor_class": "Gemma3Processor",
+  "resample": 2,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "height": 896,
+    "width": 896
+  }
+}

checkpoint-1795/processor_config.json

@@ -0,0 +1,4 @@
+{
+  "image_seq_length": 256,
+  "processor_class": "Gemma3Processor"
+}

checkpoint-1795/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:250560ab3d528161ab3659b120def6e4a9ab4b457e3399603bbcfa40db3efc90
+size 14645

checkpoint-1795/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1eb2e719ba92d61c300d3b21bda51012661191790b1797109a6a81ec3f6ab06
+size 1465

checkpoint-1795/special_tokens_map.json

@@ -0,0 +1,33 @@
+{
+  "boi_token": "<start_of_image>",
+  "bos_token": {
+    "content": "<bos>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eoi_token": "<end_of_image>",
+  "eos_token": {
+    "content": "<end_of_turn>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "image_token": "<image_soft_token>",
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}