Upload 13 files

README.md

@@ -0,0 +1,62 @@
+---
+license: other
+library_name: peft
+tags:
+- llama-factory
+- lora
+- generated_from_trainer
+base_model: IlyaGusev/saiga_mistral_7b_merged
+model-index:
+- name: 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. -->
+
+# sft
+
+This model is a fine-tuned version of [IlyaGusev/saiga_mistral_7b_merged](https://huggingface.co/IlyaGusev/saiga_mistral_7b_merged) on the gazeta 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: 5e-05
+- train_batch_size: 4
+- eval_batch_size: 2
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 2
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 16
+- total_eval_batch_size: 4
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- num_epochs: 2.0
+- mixed_precision_training: Native AMP
+
+### Training results
+
+
+
+### Framework versions
+
+- PEFT 0.10.1.dev0
+- Transformers 4.39.1
+- Pytorch 2.1.0+cu121
+- Datasets 2.18.0
+- Tokenizers 0.15.2

adapter_config.json

@@ -0,0 +1,34 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "IlyaGusev/saiga_mistral_7b_merged",
+  "bias": "none",
+  "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_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": "unsloth",
+  "target_modules": [
+    "down_proj",
+    "up_proj",
+    "v_proj",
+    "k_proj",
+    "o_proj",
+    "q_proj",
+    "gate_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04ddcb00e4ee4afb1fb1da889d9d882777a99f774f098dc8b6293ad081aa530a
+size 83945296

added_tokens.json

@@ -0,0 +1,4 @@
+{
+  "<|im_end|>": 32000,
+  "<|im_start|>": 32001
+}

all_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 2.0,
+    "train_loss": 0.9777891387939454,
+    "train_runtime": 5812.3304,
+    "train_samples_per_second": 0.688,
+    "train_steps_per_second": 0.043
+}

special_tokens_map.json

@@ -0,0 +1,35 @@
+{
+  "additional_special_tokens": [
+    "<unk>",
+    "<s>",
+    "</s>"
+  ],
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
+size 493443

tokenizer_config.json

@@ -0,0 +1,66 @@
+{
+  "add_bos_token": false,
+  "add_eos_token": false,
+  "add_prefix_space": true,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32000": {
+      "content": "<|im_end|>",
+      "lstrip": true,
+      "normalized": true,
+      "rstrip": true,
+      "single_word": false,
+      "special": false
+    },
+    "32001": {
+      "content": "<|im_start|>",
+      "lstrip": true,
+      "normalized": true,
+      "rstrip": true,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<unk>",
+    "<s>",
+    "</s>"
+  ],
+  "bos_token": "<s>",
+  "chat_template": "{% if messages[0]['role'] == 'system' %}{% set system_message = messages[0]['content'] %}{% endif %}{% if system_message is defined %}{{ system_message + '\\n' }}{% endif %}{% for message in messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ 'Human: ' + content + '\\nAssistant: ' }}{% elif message['role'] == 'assistant' %}{{ content + '</s>' + '\\n' }}{% endif %}{% endfor %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "legacy": true,
+  "model_max_length": 32768,
+  "pad_token": "<unk>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "split_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": true
+}

train_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 2.0,
+    "train_loss": 0.9777891387939454,
+    "train_runtime": 5812.3304,
+    "train_samples_per_second": 0.688,
+    "train_steps_per_second": 0.043
+}

trainer_log.jsonl

@@ -0,0 +1,26 @@
+{"current_steps": 10, "total_steps": 250, "loss": 1.3192, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.980286753286195e-05, "epoch": 0.08, "percentage": 4.0, "elapsed_time": "0:03:57", "remaining_time": "1:34:58"}
+{"current_steps": 20, "total_steps": 250, "loss": 1.2281, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.9214579028215776e-05, "epoch": 0.16, "percentage": 8.0, "elapsed_time": "0:07:50", "remaining_time": "1:30:06"}
+{"current_steps": 30, "total_steps": 250, "loss": 1.1693, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.8244412147206284e-05, "epoch": 0.24, "percentage": 12.0, "elapsed_time": "0:11:41", "remaining_time": "1:25:42"}
+{"current_steps": 40, "total_steps": 250, "loss": 1.1378, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.690766700109659e-05, "epoch": 0.32, "percentage": 16.0, "elapsed_time": "0:15:32", "remaining_time": "1:21:33"}
+{"current_steps": 50, "total_steps": 250, "loss": 1.1255, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.522542485937369e-05, "epoch": 0.4, "percentage": 20.0, "elapsed_time": "0:19:23", "remaining_time": "1:17:33"}
+{"current_steps": 60, "total_steps": 250, "loss": 1.077, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.3224215685535294e-05, "epoch": 0.48, "percentage": 24.0, "elapsed_time": "0:23:14", "remaining_time": "1:13:36"}
+{"current_steps": 70, "total_steps": 250, "loss": 1.1089, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.093559974371725e-05, "epoch": 0.56, "percentage": 28.0, "elapsed_time": "0:27:06", "remaining_time": "1:09:41"}
+{"current_steps": 80, "total_steps": 250, "loss": 1.0634, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.8395669874474915e-05, "epoch": 0.64, "percentage": 32.0, "elapsed_time": "0:30:57", "remaining_time": "1:05:46"}
+{"current_steps": 90, "total_steps": 250, "loss": 1.0618, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.564448228912682e-05, "epoch": 0.72, "percentage": 36.0, "elapsed_time": "0:34:48", "remaining_time": "1:01:52"}
+{"current_steps": 100, "total_steps": 250, "loss": 1.0099, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.302359024518024e-05, "epoch": 0.8, "percentage": 40.0, "elapsed_time": "0:38:38", "remaining_time": "0:57:57"}
+{"current_steps": 110, "total_steps": 250, "loss": 1.0095, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.9992749512860173e-05, "epoch": 0.88, "percentage": 44.0, "elapsed_time": "0:42:28", "remaining_time": "0:54:04"}
+{"current_steps": 120, "total_steps": 250, "loss": 1.0568, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.6883170138198323e-05, "epoch": 0.96, "percentage": 48.0, "elapsed_time": "0:46:19", "remaining_time": "0:50:11"}
+{"current_steps": 130, "total_steps": 250, "loss": 0.939, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.3743892045505764e-05, "epoch": 1.04, "percentage": 52.0, "elapsed_time": "0:50:10", "remaining_time": "0:46:19"}
+{"current_steps": 140, "total_steps": 250, "loss": 0.8439, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.0624423525618098e-05, "epoch": 1.12, "percentage": 56.0, "elapsed_time": "0:54:01", "remaining_time": "0:42:27"}
+{"current_steps": 150, "total_steps": 250, "loss": 0.8464, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.7573960460574133e-05, "epoch": 1.2, "percentage": 60.0, "elapsed_time": "0:57:52", "remaining_time": "0:38:35"}
+{"current_steps": 160, "total_steps": 250, "loss": 0.8446, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.4640610475167898e-05, "epoch": 1.28, "percentage": 64.0, "elapsed_time": "1:01:43", "remaining_time": "0:34:43"}
+{"current_steps": 170, "total_steps": 250, "loss": 0.8545, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.1870634250967605e-05, "epoch": 1.36, "percentage": 68.0, "elapsed_time": "1:05:35", "remaining_time": "0:30:51"}
+{"current_steps": 180, "total_steps": 250, "loss": 0.8518, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.553509672741645e-06, "epoch": 1.44, "percentage": 72.0, "elapsed_time": "1:09:25", "remaining_time": "0:27:00"}
+{"current_steps": 190, "total_steps": 250, "loss": 0.8015, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.211608069767867e-06, "epoch": 1.52, "percentage": 76.0, "elapsed_time": "1:13:16", "remaining_time": "0:23:08"}
+{"current_steps": 200, "total_steps": 250, "loss": 0.8429, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.150240033804116e-06, "epoch": 1.6, "percentage": 80.0, "elapsed_time": "1:17:07", "remaining_time": "0:19:16"}
+{"current_steps": 210, "total_steps": 250, "loss": 0.8786, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.4019145701791184e-06, "epoch": 1.68, "percentage": 84.0, "elapsed_time": "1:20:58", "remaining_time": "0:15:25"}
+{"current_steps": 220, "total_steps": 250, "loss": 0.8469, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.9942038158532407e-06, "epoch": 1.76, "percentage": 88.0, "elapsed_time": "1:24:49", "remaining_time": "0:11:34"}
+{"current_steps": 230, "total_steps": 250, "loss": 0.8525, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.493082103478517e-07, "epoch": 1.84, "percentage": 92.0, "elapsed_time": "1:28:39", "remaining_time": "0:07:42"}
+{"current_steps": 240, "total_steps": 250, "loss": 0.8663, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.8370638155215123e-07, "epoch": 1.92, "percentage": 96.0, "elapsed_time": "1:32:30", "remaining_time": "0:03:51"}
+{"current_steps": 250, "total_steps": 250, "loss": 0.8084, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.895267917501504e-09, "epoch": 2.0, "percentage": 100.0, "elapsed_time": "1:36:21", "remaining_time": "0:00:00"}
+{"current_steps": 250, "total_steps": 250, "loss": null, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 2.0, "percentage": 100.0, "elapsed_time": "1:36:21", "remaining_time": "0:00:00"}

trainer_state.json

@@ -0,0 +1,205 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.27207350730896,
+      "learning_rate": 4.980286753286195e-05,
+      "loss": 1.3192,
+      "step": 10
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.096735954284668,
+      "learning_rate": 4.9214579028215776e-05,
+      "loss": 1.2281,
+      "step": 20
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3264734745025635,
+      "learning_rate": 4.8244412147206284e-05,
+      "loss": 1.1693,
+      "step": 30
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9271645545959473,
+      "learning_rate": 4.690766700109659e-05,
+      "loss": 1.1378,
+      "step": 40
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3592920303344727,
+      "learning_rate": 4.522542485937369e-05,
+      "loss": 1.1255,
+      "step": 50
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8506274223327637,
+      "learning_rate": 4.3224215685535294e-05,
+      "loss": 1.077,
+      "step": 60
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.32490873336792,
+      "learning_rate": 4.093559974371725e-05,
+      "loss": 1.1089,
+      "step": 70
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3261430263519287,
+      "learning_rate": 3.8395669874474915e-05,
+      "loss": 1.0634,
+      "step": 80
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7234225273132324,
+      "learning_rate": 3.564448228912682e-05,
+      "loss": 1.0618,
+      "step": 90
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.270724296569824,
+      "learning_rate": 3.302359024518024e-05,
+      "loss": 1.0099,
+      "step": 100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.833041191101074,
+      "learning_rate": 2.9992749512860173e-05,
+      "loss": 1.0095,
+      "step": 110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.331037998199463,
+      "learning_rate": 2.6883170138198323e-05,
+      "loss": 1.0568,
+      "step": 120
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0229177474975586,
+      "learning_rate": 2.3743892045505764e-05,
+      "loss": 0.939,
+      "step": 130
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.161649227142334,
+      "learning_rate": 2.0624423525618098e-05,
+      "loss": 0.8439,
+      "step": 140
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.71807861328125,
+      "learning_rate": 1.7573960460574133e-05,
+      "loss": 0.8464,
+      "step": 150
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.480152130126953,
+      "learning_rate": 1.4640610475167898e-05,
+      "loss": 0.8446,
+      "step": 160
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.507754325866699,
+      "learning_rate": 1.1870634250967605e-05,
+      "loss": 0.8545,
+      "step": 170
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.9544825553894043,
+      "learning_rate": 9.553509672741645e-06,
+      "loss": 0.8518,
+      "step": 180
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.6840856075286865,
+      "learning_rate": 7.211608069767867e-06,
+      "loss": 0.8015,
+      "step": 190
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.7335338592529297,
+      "learning_rate": 5.150240033804116e-06,
+      "loss": 0.8429,
+      "step": 200
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.8081326484680176,
+      "learning_rate": 3.4019145701791184e-06,
+      "loss": 0.8786,
+      "step": 210
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.5944387912750244,
+      "learning_rate": 1.9942038158532407e-06,
+      "loss": 0.8469,
+      "step": 220
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.7935092449188232,
+      "learning_rate": 9.493082103478517e-07,
+      "loss": 0.8525,
+      "step": 230
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.1116323471069336,
+      "learning_rate": 2.8370638155215123e-07,
+      "loss": 0.8663,
+      "step": 240
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.7518155574798584,
+      "learning_rate": 7.895267917501504e-09,
+      "loss": 0.8084,
+      "step": 250
+    },
+    {
+      "epoch": 2.0,
+      "step": 250,
+      "total_flos": 1.752671732105216e+17,
+      "train_loss": 0.9777891387939454,
+      "train_runtime": 5812.3304,
+      "train_samples_per_second": 0.688,
+      "train_steps_per_second": 0.043
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 250,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "total_flos": 1.752671732105216e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:125cca0aac5b5fae0715aad2f9e93142c169bab9bedea259e82ce19e1f161487
+size 5112