Improve language tag

README.md

@@ -1,155 +1,169 @@
----
-library_name: peft
-license: apache-2.0
-base_model: Qwen/Qwen2.5-32B-Instruct
-tags:
-- generated_from_trainer
-datasets:
-- Fizzarolli/inkmix-v2
-model-index:
-- name: ckpts
-  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. -->
-
-[<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/axolotl-ai-cloud/axolotl)
-<details><summary>See axolotl config</summary>
-
-axolotl version: `0.6.0`
-```yaml
-base_model: Qwen/Qwen2.5-32B-Instruct
-
-load_in_8bit: true
-load_in_4bit: false
-
-plugins:
-  - axolotl.integrations.liger.LigerPlugin
-liger_rope: true
-liger_rms_norm: true
-liger_glu_activation: true
-liger_fused_linear_cross_entropy: true
-
-#unsloth_lora_mlp: true
-#unsloth_lora_qkv: true
-#unsloth_lora_o: true
-
-strict: false
-
-adapter: lora
-lora_r: 16
-lora_alpha: 32
-lora_dropout: 0.25
-lora_target_linear: true
-peft_layers_to_transform:
-loraplus_lr_ratio: 16
-
-chat_template: chatml
-datasets:
-  - path: Fizzarolli/inkmix-v2
-    type: chat_template
-    chat_template: tokenizer_default
-    split: train
-    field_messages: conversations
-    message_field_role: from
-    message_field_content: value
-
-dataset_prepared_path: last_run_prepared
-#val_set_size: 0.02
-output_dir: ./ckpts
-
-sequence_len: 8192
-sample_packing: true
-pad_to_sequence_len: true
-
-#wandb_project: teleut-7b-rp
-#wandb_entity:
-#wandb_watch:
-#wandb_name:
-#wandb_log_model: checkpoint
-
-# mlflow configuration if you're using it
-mlflow_tracking_uri: https://public-tracking.mlflow-e00zzfjq11ky6jcgtv.backbone-e00bgn6e63256prmhq.msp.eu-north1.nebius.cloud
-mlflow_experiment_name: tq-32b-rp-inkmixv2
-mlflow_run_name: v1
-hf_mlflow_log_artifacts: true
-
-gradient_accumulation_steps: 2
-micro_batch_size: 8
-num_epochs: 2
-optimizer: paged_adamw_8bit
-lr_scheduler: cosine
-learning_rate: 6e-5
-
-train_on_inputs: false
-group_by_length: false
-bf16: auto
-fp16:
-tf32: false
-
-gradient_checkpointing: unsloth
-gradient_checkpointing_kwargs:
-  use_reentrant: false
-early_stopping_patience:
-resume_from_checkpoint:
-logging_steps: 1
-xformers_attention:
-flash_attention: true
-
-#deepspeed: deepspeed_configs/zero3_bf16.json
-
-warmup_steps: 25
-#evals_per_epoch: 4
-eval_table_size:
-saves_per_epoch: 10
-debug:
-weight_decay: 0.05
-
-```
-
-</details><br>
-
-# ckpts
-
-This model is a fine-tuned version of [Qwen/Qwen2.5-32B-Instruct](https://huggingface.co/Qwen/Qwen2.5-32B-Instruct) on the Fizzarolli/inkmix-v2 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: 6e-05
-- train_batch_size: 8
-- eval_batch_size: 8
-- seed: 42
-- gradient_accumulation_steps: 2
-- total_train_batch_size: 16
-- optimizer: Use paged_adamw_8bit with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
-- lr_scheduler_type: cosine
-- lr_scheduler_warmup_steps: 25
-- num_epochs: 2
-
-### Training results
-
-
-
-### Framework versions
-
-- PEFT 0.14.0
-- Transformers 4.47.1
-- Pytorch 2.5.1+cu124
-- Datasets 3.1.0
+---
+library_name: peft
+license: apache-2.0
+base_model: Qwen/Qwen2.5-32B-Instruct
+tags:
+- generated_from_trainer
+datasets:
+- Fizzarolli/inkmix-v2
+language:
+- zho
+- eng
+- fra
+- spa
+- por
+- deu
+- ita
+- rus
+- jpn
+- kor
+- vie
+- tha
+- ara
+model-index:
+- name: ckpts
+  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. -->
+
+[<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/axolotl-ai-cloud/axolotl)
+<details><summary>See axolotl config</summary>
+
+axolotl version: `0.6.0`
+```yaml
+base_model: Qwen/Qwen2.5-32B-Instruct
+
+load_in_8bit: true
+load_in_4bit: false
+
+plugins:
+  - axolotl.integrations.liger.LigerPlugin
+liger_rope: true
+liger_rms_norm: true
+liger_glu_activation: true
+liger_fused_linear_cross_entropy: true
+
+#unsloth_lora_mlp: true
+#unsloth_lora_qkv: true
+#unsloth_lora_o: true
+
+strict: false
+
+adapter: lora
+lora_r: 16
+lora_alpha: 32
+lora_dropout: 0.25
+lora_target_linear: true
+peft_layers_to_transform:
+loraplus_lr_ratio: 16
+
+chat_template: chatml
+datasets:
+  - path: Fizzarolli/inkmix-v2
+    type: chat_template
+    chat_template: tokenizer_default
+    split: train
+    field_messages: conversations
+    message_field_role: from
+    message_field_content: value
+
+dataset_prepared_path: last_run_prepared
+#val_set_size: 0.02
+output_dir: ./ckpts
+
+sequence_len: 8192
+sample_packing: true
+pad_to_sequence_len: true
+
+#wandb_project: teleut-7b-rp
+#wandb_entity:
+#wandb_watch:
+#wandb_name:
+#wandb_log_model: checkpoint
+
+# mlflow configuration if you're using it
+mlflow_tracking_uri: https://public-tracking.mlflow-e00zzfjq11ky6jcgtv.backbone-e00bgn6e63256prmhq.msp.eu-north1.nebius.cloud
+mlflow_experiment_name: tq-32b-rp-inkmixv2
+mlflow_run_name: v1
+hf_mlflow_log_artifacts: true
+
+gradient_accumulation_steps: 2
+micro_batch_size: 8
+num_epochs: 2
+optimizer: paged_adamw_8bit
+lr_scheduler: cosine
+learning_rate: 6e-5
+
+train_on_inputs: false
+group_by_length: false
+bf16: auto
+fp16:
+tf32: false
+
+gradient_checkpointing: unsloth
+gradient_checkpointing_kwargs:
+  use_reentrant: false
+early_stopping_patience:
+resume_from_checkpoint:
+logging_steps: 1
+xformers_attention:
+flash_attention: true
+
+#deepspeed: deepspeed_configs/zero3_bf16.json
+
+warmup_steps: 25
+#evals_per_epoch: 4
+eval_table_size:
+saves_per_epoch: 10
+debug:
+weight_decay: 0.05
+
+```
+
+</details><br>
+
+# ckpts
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-32B-Instruct](https://huggingface.co/Qwen/Qwen2.5-32B-Instruct) on the Fizzarolli/inkmix-v2 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: 6e-05
+- train_batch_size: 8
+- eval_batch_size: 8
+- seed: 42
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 16
+- optimizer: Use paged_adamw_8bit with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_steps: 25
+- num_epochs: 2
+
+### Training results
+
+
+
+### Framework versions
+
+- PEFT 0.14.0
+- Transformers 4.47.1
+- Pytorch 2.5.1+cu124
+- Datasets 3.1.0
 - Tokenizers 0.21.0