Create README.md

README.md

@@ -0,0 +1,190 @@
+---
+library_name: transformers
+base_model:
+- Qwen/Qwen3-8B
+---
+
+<!-- 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: `0.13.0.dev0`
+```yaml
+base_model: /mnt/qwen3-8b
+
+load_in_8bit: false
+load_in_4bit: false
+strict: false
+
+plugins:
+  - axolotl.integrations.liger.LigerPlugin
+
+liger_rope: true
+liger_rms_norm: true
+liger_glu_activation: true
+liger_layer_norm: true
+liger_fused_linear_cross_entropy: true
+
+chat_template: qwen3
+datasets:
+  # Dolci
+  - path: /home/aithucchien/Unicorn/data/dolci_vi_5k_single.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+
+  # 12 Task
+  - path: /home/aithucchien/luannd/Unicorn/data/synthetic/output_12_task_split/task2_sua_loi_sai_answer.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+  - path: /home/aithucchien/luannd/Unicorn/data/synthetic/output_12_task_split/task3_goi_mo_y_tuong_answer.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+  - path: /home/aithucchien/luannd/Unicorn/data/synthetic/output_12_task_split/task10_hoc_tap_tuong_tac_answer.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+  - path: /home/aithucchien/luannd/Unicorn/data/synthetic/output_12_task_split/task1_tra_loi_cau_hoi_extended_2.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+  - path: /home/aithucchien/luannd/Unicorn/data/synthetic/output_12_task_split/task8_tao_tai_lieu_giang_day_answer.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+  - path: /home/aithucchien/luannd/Unicorn/data/synthetic/output_12_task_split/task9_tao_noi_dung_ca_nhan_hoa_answer.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+  - path: /home/aithucchien/luannd/Unicorn/data/synthetic/output_12_task_split/task5_ho_tro_tam_ly_answer.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+  - path: /home/aithucchien/luannd/Unicorn/data/synthetic/output_12_task_split/task7_cham_diem_tu_dong_answer.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+  - path: /home/aithucchien/luannd/Unicorn/data/synthetic/output_12_task_split/task6_tao_bo_cau_hoi_answer.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+  - path: /home/aithucchien/luannd/Unicorn/data/synthetic/output_12_task_split/task1_tra_loi_cau_hoi_extended.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+  - path: /home/aithucchien/luannd/Unicorn/data/synthetic/output_12_task_split/task4_hoc_tap_ca_nhan_hoa_answer.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+  - path: /home/aithucchien/luannd/Unicorn/data/synthetic/output_12_task_split/task1_tra_loi_cau_hoi_answer.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+  - path: /home/aithucchien/luannd/Unicorn/data/synthetic/output_12_task_split/task11_tu_van_an_toan_answer.jsonl
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+
+
+output_dir: ./outputs/qwen3_8b_dolci_vi_5k_single_12task/
+
+sequence_len: 32768
+sample_packing: true
+flex_attention: true
+
+
+flex_attn_compile_kwargs:
+  dynamic: false
+  mode: max-autotune-no-cudagraphs
+
+wandb_project: aithucchien
+wandb_entity:
+wandb_watch:
+wandb_name: qwen3_8b_dolci_vi_5k_single_12task
+wandb_log_model:
+
+gradient_accumulation_steps: 4
+micro_batch_size: 1
+num_epochs: 3
+optimizer: adamw_torch_fused
+lr_scheduler: cosine
+learning_rate: 2e-5
+
+bf16: true
+tf32: true
+
+resume_from_checkpoint:
+logging_steps: 1
+
+saves_per_epoch: 1
+
+warmup_ratio: 0.1
+weight_decay: 0.0
+fsdp:
+  - full_shard
+  - auto_wrap
+
+fsdp_config:
+  fsdp_version: 2
+  fsdp_offload_params: false
+  fsdp_cpu_ram_efficient_loading: true
+  fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
+  fsdp_transformer_layer_cls_to_wrap: Qwen3DecoderLayer
+  fsdp_state_dict_type: FULL_STATE_DICT
+  fsdp_sharding_strategy: FULL_SHARD
+  fsdp_reshard_after_forward: true
+  fsdp_activation_checkpointing: true
+
+special_tokens:
+
+```
+
+</details><br>
+
+# Unicorn-R3
+
+## 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: 2e-05
+- train_batch_size: 1
+- eval_batch_size: 1
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 2
+- gradient_accumulation_steps: 4
+- total_train_batch_size: 8
+- total_eval_batch_size: 2
+- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 18
+- training_steps: 186
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 4.57.3
+- Pytorch 2.8.0+cu128
+- Datasets 4.4.1
+- Tokenizers 0.22.1