Skip to content

[LLM] Add DeepseekV3 #9738

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ZHUI merged 9 commits into from
Jan 10, 2025
Merged

[LLM] Add DeepseekV3 #9738

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
bcdcab2
update deepseek-v2
DrownFish19 Dec 31, 2024
4a2be33
add deepseek_v3
DrownFish19 Jan 3, 2025
99c4def
Merge branch 'PaddlePaddle:develop' into dev_20241231_add_deepseekv3
DrownFish19 Jan 7, 2025
8d11a23
update for deepseekv3
DrownFish19 Jan 7, 2025
d89d6d4
update prepare_inputs_for_generation
DrownFish19 Jan 7, 2025
20bd1ea
update for predict
DrownFish19 Jan 7, 2025
f9abe9c
update TP and model load
DrownFish19 Jan 7, 2025
cac02e4
Merge branch 'dev_20241231_add_deepseekv3' of github.com:DrownFish19/…
DrownFish19 Jan 8, 2025
4d61571
update deepseekv3 model_ids
DrownFish19 Jan 10, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
27 changes: 14 additions & 13 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -90,6 +90,7 @@
| [ChatGLM2](https://github.com/PaddlePaddle/PaddleNLP/tree/develop/llm/config/chatglm2) | THUDM/chatglm2-6b |
| [ChatGLM3](https://github.com/PaddlePaddle/PaddleNLP/tree/develop/llm/config/chatglm2) | THUDM/chatglm3-6b |
| [DeepSeekV2](https://github.com/PaddlePaddle/PaddleNLP/blob/develop/llm/config/deepseek-v2) | deepseek-ai/DeepSeek-V2, deepseek-ai/DeepSeek-V2-Chat, deepseek-ai/DeepSeek-V2-Lite, deepseek-ai/DeepSeek-V2-Lite-Chat, deepseek-ai/DeepSeek-Coder-V2-Base, deepseek-ai/DeepSeek-Coder-V2-Instruct, deepseek-ai/DeepSeek-Coder-V2-Lite-Base, deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct |
| [DeepSeekV3](https://github.com/PaddlePaddle/PaddleNLP/blob/develop/llm/config/deepseek-v2) | deepseek-ai/DeepSeek-V3, deepseek-ai/DeepSeek-V3-Base |
| [Gemma](https://github.com/PaddlePaddle/PaddleNLP/tree/develop/llm/config/gemma) | google/gemma-7b, google/gemma-7b-it, google/gemma-2b, google/gemma-2b-it |
| [Mistral](https://github.com/PaddlePaddle/PaddleNLP/tree/develop/llm/config/mistral) | mistralai/Mistral-7B-Instruct-v0.3, mistralai/Mistral-7B-v0.1 |
| [Mixtral](https://github.com/PaddlePaddle/PaddleNLP/tree/develop/llm/config/mixtral) | mistralai/Mixtral-8x7B-Instruct-v0.1 |
Expand Down Expand Up @@ -130,19 +131,19 @@


| Model | Pretrain | SFT | LoRA | FlashMask | Prefix Tuning | DPO/SimPO/ORPO/KTO | RLHF | Mergekit | Quantization |
|--------------------------------------------|:--------:|:---:|:----:|:---------:|:-------------:|:--------------:|:----:|:-----:|:------------:|
| [Llama](./llm/config/llama) | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
| [Qwen](./llm/config/qwen) | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | 🚧 | ✅ | 🚧 |
| [Mixtral](./llm/config/mixtral) | ✅ | ✅ | ✅ | 🚧 | 🚧 | ✅ | 🚧 | ✅ | 🚧 |
| [Mistral](./llm/config/mistral) | ✅ | ✅ | ✅ | 🚧 | ✅ | ✅ | 🚧 | ✅ | 🚧 |
| [Baichuan/Baichuan2](./llm/config/llama) | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | 🚧 | ✅ | ✅ |
| [ChatGLM-6B](./llm/config/chatglm) | ✅ | ✅ | ✅ | 🚧 | ✅ | 🚧 | 🚧 | ✅ | ✅ |
| [ChatGLM2/ChatGLM3](./llm/config/chatglm2) | ✅ | ✅ | ✅ | 🚧 | ✅ | ✅ | 🚧 | ✅ | ✅ |
| [Bloom](./llm/config/bloom) | ✅ | ✅ | ✅ | 🚧 | ✅ | 🚧 | 🚧 | ✅ | ✅ |
| [GPT-3](./llm/config/gpt-3) | ✅ | ✅ | 🚧 | 🚧 | 🚧 | 🚧 | 🚧 | ✅ | 🚧 |
| [OPT](./llm/config/opt) | ✅ | ✅ | ✅ | 🚧 | 🚧 | 🚧 | 🚧 | ✅ | 🚧 |
| [Gemma](./llm/config/gemma) | ✅ | ✅ | ✅ | 🚧 | 🚧 | ✅ | 🚧 | ✅ | 🚧 |
| [Yuan](./llm/config/yuan) | ✅ | ✅ | ✅ | 🚧 | 🚧 | ✅ | 🚧 | ✅ | 🚧 |
|--------------------------------------------|:--------:|:---:|:----:|:---------:|:-------------:|:------------------:|:----:|:--------:|:------------:|
| [Llama](./llm/config/llama) | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
| [Qwen](./llm/config/qwen) | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | 🚧 | ✅ | 🚧 |
| [Mixtral](./llm/config/mixtral) | ✅ | ✅ | ✅ | 🚧 | 🚧 | ✅ | 🚧 | ✅ | 🚧 |
| [Mistral](./llm/config/mistral) | ✅ | ✅ | ✅ | 🚧 | ✅ | ✅ | 🚧 | ✅ | 🚧 |
| [Baichuan/Baichuan2](./llm/config/llama) | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | 🚧 | ✅ | ✅ |
| [ChatGLM-6B](./llm/config/chatglm) | ✅ | ✅ | ✅ | 🚧 | ✅ | 🚧 | 🚧 | ✅ | ✅ |
| [ChatGLM2/ChatGLM3](./llm/config/chatglm2) | ✅ | ✅ | ✅ | 🚧 | ✅ | ✅ | 🚧 | ✅ | ✅ |
| [Bloom](./llm/config/bloom) | ✅ | ✅ | ✅ | 🚧 | ✅ | 🚧 | 🚧 | ✅ | ✅ |
| [GPT-3](./llm/config/gpt-3) | ✅ | ✅ | 🚧 | 🚧 | 🚧 | 🚧 | 🚧 | ✅ | 🚧 |
| [OPT](./llm/config/opt) | ✅ | ✅ | ✅ | 🚧 | 🚧 | 🚧 | 🚧 | ✅ | 🚧 |
| [Gemma](./llm/config/gemma) | ✅ | ✅ | ✅ | 🚧 | 🚧 | ✅ | 🚧 | ✅ | 🚧 |
| [Yuan](./llm/config/yuan) | ✅ | ✅ | ✅ | 🚧 | 🚧 | ✅ | 🚧 | ✅ | 🚧 |
* [大模型推理](./llm/docs/predict/inference.md)已支持 LLaMA 系列、Qwen 系列、Mistral 系列、ChatGLM 系列、Bloom 系列和 Baichuan 系列，支持 Weight Only INT8及 INT4推理，支持 WAC（权重、激活、Cache KV）进行 INT8、FP8量化的推理，【LLM】模型推理支持列表如下：

| 模型名称/量化类型支持 | FP16/BF16 | WINT8 | WINT4 | INT8-A8W8 | FP8-A8W8 | INT8-A8W8C8 |
Expand Down
2 changes: 1 addition & 1 deletion paddlenlp/generation/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -742,7 +742,7 @@ def generate(
# ['是的', '嗯嗯']
"""
if generation_config is None:
if self.generation_config._from_model_config:
if self.generation_config is None or self.generation_config._from_model_config:
new_generation_config = GenerationConfig.from_model_config(self.config)
if new_generation_config != self.generation_config:
logger.warning(
Expand Down
1 change: 1 addition & 0 deletions paddlenlp/quantization/quantization_config.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -52,6 +52,7 @@ def __init__(
weight_double_quant_block_size=256,
weight_quant_method="abs_max_channel_wise",
act_quant_method="abs_max",
**kwargs,
):
if weight_quantize_algo is not None and weight_quantize_algo not in [
"weight_only_int8",
Expand Down
Loading