Skip to content

Commit 24fa34c

Browse files
authored
Fix README.md (#4882)
* Revise example prompt and answer to English Updated example prompt and answer in README with English content. * Fix typo issue and add link to generative ai readme.md
1 parent 16876be commit 24fa34c

2 files changed

Lines changed: 11 additions & 9 deletions

File tree

‎ generative_ai/README.md‎

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -6,7 +6,8 @@ These examples showcases Amazon SageMaker's capabilities in the exciting field o
66

77
- [Fine-tuning and deploying a Hugging Face summarization model on SageMaker with your own scripts and dataset](sm-finetuning_huggingface_with_your_own_scripts_and_data/sm-finetuning_huggingface_with_your_own_scripts_and_data.ipynb)
88
- [Fine-tuning and deploying the Mixtral 8x7B LLM In SageMaker with Hugging Face, using QLoRA Parameter-Efficient Fine-Tuning](sm-mixtral_8x7b_fine_tune_and_deploy/sm-mixtral_8x7b_fine_tune_and_deploy.ipynb)
9-
- [Serve large models on SageMaker with DeepSpeed Container](sm-djl_deepspeed_bloom_176b_deploy.ipynb)
9+
- [Qwen 8B LLM Fine-tuning with SFT and GRPO, and Deployment on AWS SageMaker](sm-qwen3_8b_fine_tune_and_deploy/README.md)
10+
- [Serve large models on SageMaker with DeepSpeed Container](sm-mixtral_8x7b_fine_tune_and_deploy/sm-mixtral_8x7b_fine_tune_and_deploy.ipynb)(sm-djl_deepspeed_bloom_176b_deploy.ipynb)
1011
- [Accelerate SageMaker-PyTorch FSDP Training of Llama-v2 (or GPT-NeoX) with FP8 on P5 instances](sm-fsdp_training_of_llama_v2_with_fp8_on_p5.ipynb)
1112
- [Fine-tune Code Llama, Deploy and Evaluate the Fine-tuning with Human-eval Repository](sm-jumpstart_foundation_code_llama_fine_tuning_human_eval.ipynb)
1213
- [SageMaker JumpStart Foundation Models - Fine-tuning text generation GPT-J 6B model on domain specific dataset](sm-jumpstart_foundation_finetuning_gpt_j_6b_domain_adaptation.ipynb)

generative_ai/sm-qwen3_8b_fine_tune_and_deploy/README.md renamed to generative_ai/sm-qwen3_8b_fine_tune_and_deploy/README.md

Lines changed: 9 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -626,21 +626,21 @@ The reward function in `reward_function/math.py` computes a multi-metric score f
626626
627627
| Metric | Weight | Description |
628628
|--------|--------|-------------|
629-
| `recall` | 35% | Proportion of ground truth tags found |
630-
| `accuracy` | 35% | Classification accuracy of matched tags |
631-
| `precision` | 20% | Penalizes extra predicted tags |
632-
| `match_quality` | 5% | Fuzzy matching quality score |
629+
| `recall` | 30% | Proportion of ground truth tags found |
630+
| `accuracy` | 30% | Classification accuracy of matched tags |
631+
| `precision` | 25% | Penalizes extra predicted tags |
632+
| `match_quality` | 10% | Fuzzy matching quality score |
633633
| `formatting` | 5% | Correct output format (9 categories) |
634634
635635
To customize the reward function, edit `2_trainning_grpo/docker/reward_function/math.py`:
636636
637637
```python
638638
def compute_score(
639639
reward_inputs: list[dict[str, Any]],
640-
recall_weight: float = 0.35,
641-
precision_weight: float = 0.2,
642-
accuracy_weight: float = 0.35,
643-
match_quality_weight: float = 0.05,
640+
recall_weight: float = 0.30,
641+
precision_weight: float = 0.25,
642+
accuracy_weight: float = 0.30,
643+
match_quality_weight: float = 0.10,
644644
formatting_weight: float = 0.05
645645
) -> list[dict[str, float]]:
646646
"""
@@ -683,6 +683,7 @@ GRPO training uses **Parquet format** with prompt-answer pairs. Upload your data
683683
| `answer` | string | The expected ground truth response (used by reward function) |
684684
685685
**Example row:**
686+
686687
| Column | Content |
687688
|--------|---------|
688689
| `problem` | You are a professional and rigorous product tagging expert, responsible for automatically generating and classifying tags based on the provided product information...<br><br>Product Name: MI Amazon Usb Type-C Cable Smartphone Charging (Black) \|Connectivity: Usb 2.0 (Sync And Charging)\| Universal For All Type-C Devices (Grey)... |

0 commit comments

Comments
 (0)