Hugging Face's logo

tencent
/
Hy3-preview-Base

Text Generation
Transformers
Safetensors
hy_v3
conversational
Model card Files
xet
 Community
2
yiqichen01 commited on
Commit
eef4b1c
·
verified ·
1 Parent(s): 5f27182

Upload folder using huggingface_hub

Browse files
Files changed (2) hide show
  1. README.md +6 -10
  2. README_CN.md +6 -6
README.md CHANGED
@@ -1,9 +1,5 @@
1
- ---
2
- license: other
3
- library_name: transformers
4
- ---
5
  <p align="left">
6
- <a href="https://huggingface.co/tencent/Hy3-preview-Base/blob/main/README_CN.md">中文</a>&nbsp;｜&nbsp;English
7
  </p>
8
  <br>
9
 
@@ -162,7 +158,7 @@ from openai import OpenAI
162
  client = OpenAI(base_url="http://localhost:8000/v1", api_key="EMPTY")
163
 
164
  response = client.chat.completions.create(
165
- model="hy3-preview-base",
166
  messages=[
167
  {"role": "user", "content": "Hello! Can you briefly introduce yourself?"},
168
  ],
@@ -198,14 +194,14 @@ uv pip install --editable . --torch-backend=auto
198
  Start the vLLM server with MTP enabled:
199
 
200
  ```bash
201
- vllm serve tencent/Hy3-preview-base \
202
  --tensor-parallel-size 8 \
203
  --speculative-config.method mtp \
204
  --speculative-config.num_speculative_tokens 1 \
205
  --tool-call-parser hy_v3 \
206
  --reasoning-parser hy_v3 \
207
  --enable-auto-tool-choice \
208
- --served-model-name hy3-preview-base
209
  ```
210
 
211
  ### SGLang
@@ -223,7 +219,7 @@ Launch SGLang server with MTP enabled:
223
 
224
  ```bash
225
  python3 -m sglang.launch_server \
226
- --model tencent/Hy3-preview-base \
227
  --tp 8 \
228
  --tool-call-parser hunyuan \
229
  --reasoning-parser hunyuan \
@@ -231,7 +227,7 @@ python3 -m sglang.launch_server \
231
  --speculative-eagle-topk 1 \
232
  --speculative-num-draft-tokens 2 \
233
  --speculative-algorithm EAGLE \
234
- --served-model-name hy3-preview-base
235
  ```
236
 
237
  ## Training
 
 
 
 
 
1
  <p align="left">
2
+ <a href="README_CN.md">中文</a>&nbsp;｜&nbsp;English
3
  </p>
4
  <br>
5
 
 
158
  client = OpenAI(base_url="http://localhost:8000/v1", api_key="EMPTY")
159
 
160
  response = client.chat.completions.create(
161
+ model="hy3-preview-Base",
162
  messages=[
163
  {"role": "user", "content": "Hello! Can you briefly introduce yourself?"},
164
  ],
 
194
  Start the vLLM server with MTP enabled:
195
 
196
  ```bash
197
+ vllm serve tencent/Hy3-preview-Base \
198
  --tensor-parallel-size 8 \
199
  --speculative-config.method mtp \
200
  --speculative-config.num_speculative_tokens 1 \
201
  --tool-call-parser hy_v3 \
202
  --reasoning-parser hy_v3 \
203
  --enable-auto-tool-choice \
204
+ --served-model-name hy3-preview-Base
205
  ```
206
 
207
  ### SGLang
 
219
 
220
  ```bash
221
  python3 -m sglang.launch_server \
222
+ --model tencent/Hy3-preview-Base \
223
  --tp 8 \
224
  --tool-call-parser hunyuan \
225
  --reasoning-parser hunyuan \
 
227
  --speculative-eagle-topk 1 \
228
  --speculative-num-draft-tokens 2 \
229
  --speculative-algorithm EAGLE \
230
+ --served-model-name hy3-preview-Base
231
  ```
232
 
233
  ## Training
README_CN.md CHANGED
@@ -1,5 +1,5 @@
1
  <p align="left">
2
- <a href="https://huggingface.co/tencent/Hy3-preview-Base">English</a>&nbsp;｜&nbsp;中文
3
  </p>
4
  <br>
5
 
@@ -157,7 +157,7 @@ from openai import OpenAI
157
  client = OpenAI(base_url="http://localhost:8000/v1", api_key="EMPTY")
158
 
159
  response = client.chat.completions.create(
160
- model="hy3-preview-base",
161
  messages=[
162
  {"role": "user", "content": "你好！请简单介绍一下你自己。"},
163
  ],
@@ -194,14 +194,14 @@ uv pip install --editable . --torch-backend=auto
194
  启动 vLLM 服务，开启 MTP：
195
 
196
  ```bash
197
- vllm serve tencent/Hy3-preview-base \
198
  --tensor-parallel-size 8 \
199
  --speculative-config.method mtp \
200
  --speculative-config.num_speculative_tokens 1 \
201
  --tool-call-parser hy_v3 \
202
  --reasoning-parser hy_v3 \
203
  --enable-auto-tool-choice \
204
- --served-model-name hy3-preview-base
205
  ```
206
 
207
  ### SGLang
@@ -220,7 +220,7 @@ pip3 install -e "python"
220
 
221
  ```bash
222
  python3 -m sglang.launch_server \
223
- --model tencent/Hy3-preview-base \
224
  --tp 8 \
225
  --tool-call-parser hunyuan \
226
  --reasoning-parser hunyuan \
@@ -228,7 +228,7 @@ python3 -m sglang.launch_server \
228
  --speculative-eagle-topk 1 \
229
  --speculative-num-draft-tokens 2 \
230
  --speculative-algorithm EAGLE \
231
- --served-model-name hy3-preview-base
232
  ```
233
 
234
  ## 模型训练
 
1
  <p align="left">
2
+ <a href="README.md">English</a>&nbsp;｜&nbsp;中文
3
  </p>
4
  <br>
5
 
 
157
  client = OpenAI(base_url="http://localhost:8000/v1", api_key="EMPTY")
158
 
159
  response = client.chat.completions.create(
160
+ model="hy3-preview-Base",
161
  messages=[
162
  {"role": "user", "content": "你好！请简单介绍一下你自己。"},
163
  ],
 
194
  启动 vLLM 服务，开启 MTP：
195
 
196
  ```bash
197
+ vllm serve tencent/Hy3-preview-Base \
198
  --tensor-parallel-size 8 \
199
  --speculative-config.method mtp \
200
  --speculative-config.num_speculative_tokens 1 \
201
  --tool-call-parser hy_v3 \
202
  --reasoning-parser hy_v3 \
203
  --enable-auto-tool-choice \
204
+ --served-model-name hy3-preview-Base
205
  ```
206
 
207
  ### SGLang
 
220
 
221
  ```bash
222
  python3 -m sglang.launch_server \
223
+ --model tencent/Hy3-preview-Base \
224
  --tp 8 \
225
  --tool-call-parser hunyuan \
226
  --reasoning-parser hunyuan \
 
228
  --speculative-eagle-topk 1 \
229
  --speculative-num-draft-tokens 2 \
230
  --speculative-algorithm EAGLE \
231
+ --served-model-name hy3-preview-Base
232
  ```
233
 
234
  ## 模型训练