Upload LoRA adapter - README.md

README.md

@@ -0,0 +1,158 @@
+---
+library_name: peft
+license: apache-2.0
+base_model: Qwen/Qwen3-4B-Instruct-2507
+tags:
+- axolotl
+- base_model:adapter:Qwen/Qwen3-4B-Instruct-2507
+- lora
+- transformers
+datasets:
+- custom
+pipeline_tag: text-generation
+model-index:
+- name: agent-0916
+  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: `0.12.2`
+```yaml
+# Automatically upload checkpoint and final model to HF
+# hub_model_id: username/custom_model_name
+# 是否以 8-bit 精度加载模型
+load_in_8bit: false
+# 是否以 4-bit 精度加载模型（与QLoRA绑定, 强制使用）
+load_in_4bit: false
+# 是否严格匹配模型结构，关闭表示可加载少部分差异结构（如以适配 adapter）
+# strict: false
+base_model: Qwen/Qwen3-4B-Instruct-2507
+# 数据集设置
+chat_template: qwen3
+datasets:
+  - path: /workspace/train_dir/agent_train_data_all_fix.json # - 表示列表（list）中的一项, 即可以同时使用多个数据集
+    type: chat_template # chat_template(自定义格式) alpaca
+    roles_to_train: ["assistant"]
+    field_messages: messages # 标识的字段
+    message_property_mappings:  # message_property_mappings={'role':'role', 'content':'content'})
+      role: role
+      content: content
+dataset_prepared_path:
+val_set_size: 0.05
+output_dir: checkpoints/0916
+sequence_len: 8192 # 模型所能处理的最大上下文长度(默认2048)
+pad_to_sequence_len: true
+# context_parallel_size: 2 # 长序列拆分至多个GPU(强制要求 mirco_batch_size: 1)
+sample_packing: false # 在训练时将多个样本拼接（packing）成一个长序列(sequence_len)输入到模型中，以提高训练效率。
+eval_sample_packing: false # 评估时拼接多个样本
+# 训练超参数
+adapter: lora  # lora qlora
+lora_model_dir:
+lora_r: 16 # lora_r默认首选 16，平衡精度与显存
+lora_alpha: 64 # 缩放系数，用于控制 LoRA 的影响力, 一般设为 2*r 或 4*r
+lora_dropout: 0.05
+lora_target_linear: true
+micro_batch_size: 8 # 微批次大小 94G的H100可以设为4(Token为1w)
+gradient_accumulation_steps: 4 # 梯度累积: 将多个微批次的梯度(micro_batch_size)累积起来，然后更新模型权重 有效 Batch 常取 16: 小于 8 训练会抖，大于 32 只会更耗时、收益有限
+auto_find_batch_size: false # 允许Axolotl不断调整batch_size  ⚠️Zero-3不适用
+num_epochs: 1
+optimizer: adamw_torch_fused
+lr_scheduler: cosine
+learning_rate: 2e-5
+# bf16: auto + tf32: true，可获得更好的稳定性和性能。
+bf16: auto
+tf32: true
+# early_stopping_patience:
+gradient_checkpointing: true
+gradient_checkpointing_kwargs:
+  use_reentrant: false
+# auto_resume_from_checkpoints: true #自动从output_dir寻找最新checkpoint断点恢复
+logging_steps: 1
+flash_attention: true
+warmup_steps: 10
+evals_per_epoch: 4
+saves_per_epoch: 1
+weight_decay: 0.0
+# deepspeed: /workspace/deepspeed_configs/zero2.json
+# fsdp:
+#   - full_shard
+#   - auto_wrap
+# fsdp_config:
+#   fsdp_limit_all_gathers: true
+#   fsdp_sync_module_states: true
+#   fsdp_offload_params: true
+#   fsdp_use_orig_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
+# special_tokens:
+# wandb_project:
+# wandb_entity:
+# wandb_watch:
+# wandb_name:
+# wandb_log_model:
+```
+
+</details><br>
+
+# checkpoints/0916
+
+This model is a fine-tuned version of [Qwen/Qwen3-4B-Instruct-2507](https://huggingface.co/Qwen/Qwen3-4B-Instruct-2507) on the /workspace/train_dir/agent_train_data_all_fix.json dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0333
+- Memory/max Mem Active(gib): 134.01
+- Memory/max Mem Allocated(gib): 134.01
+- Memory/device Mem Reserved(gib): 135.43
+
+## 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: 8
+- eval_batch_size: 8
+- seed: 42
+- gradient_accumulation_steps: 4
+- total_train_batch_size: 32
+- 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: 10
+- training_steps: 769
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss | Mem Active(gib) | Mem Allocated(gib) | Mem Reserved(gib) |
+|:-------------:|:------:|:----:|:---------------:|:---------------:|:------------------:|:-----------------:|
+| No log        | 0      | 0    | 1.4283          | 103.1           | 103.1              | 103.74            |
+| 0.0654        | 0.2510 | 193  | 0.0552          | 134.01          | 134.01             | 135.39            |
+| 0.0286        | 0.5020 | 386  | 0.0407          | 134.01          | 134.01             | 135.43            |
+| 0.0206        | 0.7529 | 579  | 0.0333          | 134.01          | 134.01             | 135.43            |
+
+
+### Framework versions
+
+- PEFT 0.17.0
+- Transformers 4.55.2
+- Pytorch 2.6.0+cu126
+- Datasets 4.0.0
+- Tokenizers 0.21.4