初始化项目，由ModelHub XC社区提供模型

Model: W-61/qwen3-8b-base-new-dpo-ultrafeedback-4xh200-batch-128-q_t-0.45-s_star-0.35-20260430-143919 Source: Original Platform
2026-05-30 15:45:24 +08:00
commit 7ca2df9d3d
501 changed files with 166907 additions and 0 deletions
--- a/README.md
+++ b/README.md
@@ -0,0 +1,81 @@
+---
+library_name: transformers
+base_model: jackf857/qwen3-8b-base-sft-ultrachat-4xh200-batch-128
+tags:
+- alignment-handbook
+- new-dpo
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
+model-index:
+- name: qwen3-8b-base-new-dpo-ultrafeedback-4xh200-batch-128-q_t-0.45-s_star-0.35-20260430-143919
+  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. -->
+
+# qwen3-8b-base-new-dpo-ultrafeedback-4xh200-batch-128-q_t-0.45-s_star-0.35-20260430-143919
+
+This model is a fine-tuned version of [jackf857/qwen3-8b-base-sft-ultrachat-4xh200-batch-128](https://huggingface.co/jackf857/qwen3-8b-base-sft-ultrachat-4xh200-batch-128) on the HuggingFaceH4/ultrafeedback_binarized dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.5890
+- Fcm Dpo/beta: 0.0056
+- Margin Dpo/margin Mean: 51.3408
+- Margin Dpo/margin Std: 80.2587
+- Logps/chosen: -323.1461
+- Logps/rejected: -358.7158
+- Logps/ref Chosen: -280.4167
+- Logps/ref Rejected: -264.6455
+- Kl/chosen Kl Mean: -42.7294
+- Kl/rejected Kl Mean: -94.0702
+- Kl/mean: -68.3998
+- Kl/std: 69.5920
+- Logits/chosen: 1.3840
+- Logits/rejected: 1.4393
+
+## 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-07
+- train_batch_size: 4
+- eval_batch_size: 2
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- gradient_accumulation_steps: 8
+- total_train_batch_size: 128
+- total_eval_batch_size: 8
+- optimizer: Use OptimizerNames.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
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss | Fcm Dpo/beta | Margin Dpo/margin Mean | Margin Dpo/margin Std | Logps/chosen | Logps/rejected | Logps/ref Chosen | Logps/ref Rejected | Kl/chosen Kl Mean | Kl/rejected Kl Mean | Kl/mean  | Kl/std  | Logits/chosen | Logits/rejected |
+|:-------------:|:------:|:----:|:---------------:|:------------:|:----------------------:|:---------------------:|:------------:|:--------------:|:----------------:|:------------------:|:-----------------:|:-------------------:|:--------:|:-------:|:-------------:|:---------------:|
+| 4.9323        | 0.4188 | 200  | 0.5859          | 0.0124       | 24.4991                | 38.5860               | -285.8270    | -294.5549      | -280.4167        | -264.6455          | -5.4103           | -29.9094            | -17.6598 | 33.7800 | 1.5348        | 1.5657          |
+| 4.617         | 0.8377 | 400  | 0.5890          | 0.0056       | 51.3408                | 80.2587               | -323.1461    | -358.7158      | -280.4167        | -264.6455          | -42.7294          | -94.0702            | -68.3998 | 69.5920 | 1.3840        | 1.4393          |
+
+
+### Framework versions
+
+- Transformers 4.51.0
+- Pytorch 2.3.1+cu121
+- Datasets 2.21.0
+- Tokenizers 0.21.4