Instructions to use katuni4ka/tiny-random-minicpm with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use katuni4ka/tiny-random-minicpm with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="katuni4ka/tiny-random-minicpm", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("katuni4ka/tiny-random-minicpm", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use katuni4ka/tiny-random-minicpm with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "katuni4ka/tiny-random-minicpm"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "katuni4ka/tiny-random-minicpm",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/katuni4ka/tiny-random-minicpm

SGLang

How to use katuni4ka/tiny-random-minicpm with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "katuni4ka/tiny-random-minicpm" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "katuni4ka/tiny-random-minicpm",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "katuni4ka/tiny-random-minicpm" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "katuni4ka/tiny-random-minicpm",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use katuni4ka/tiny-random-minicpm with Docker Model Runner:
```
docker model run hf.co/katuni4ka/tiny-random-minicpm
```

katuni4ka commited on Oct 17, 2024

Commit

edc7c55

verified ·

1 Parent(s): 85cbfcb

remove flash_attn imports and usage

Browse files

Files changed (1) hide show

modeling_minicpm.py +6 -6

modeling_minicpm.py CHANGED Viewed

@@ -51,11 +51,11 @@ from transformers.utils.import_utils import is_torch_fx_available
 from .configuration_minicpm import MiniCPMConfig
 import re
-try:
-    from flash_attn import flash_attn_func, flash_attn_varlen_func
-    from flash_attn.bert_padding import index_first_axis, pad_input, unpad_input  # noqa
-except:
-    pass
 # This makes `_prepare_4d_causal_attention_mask` a leaf function in the FX graph.
@@ -755,7 +755,7 @@ class MiniCPMSdpaAttention(MiniCPMAttention):
 MINICPM_ATTENTION_CLASSES = {
     "eager": MiniCPMAttention,
-    "flash_attention_2": MiniCPMFlashAttention2,
     "sdpa": MiniCPMSdpaAttention,
 }

 from .configuration_minicpm import MiniCPMConfig
 import re
+#try:
+#    from flash_attn import flash_attn_func, flash_attn_varlen_func
+#    from flash_attn.bert_padding import index_first_axis, pad_input, unpad_input  # noqa
+#except:
+#    pass
 # This makes `_prepare_4d_causal_attention_mask` a leaf function in the FX graph.
 MINICPM_ATTENTION_CLASSES = {
     "eager": MiniCPMAttention,
+   #"flash_attention_2": MiniCPMFlashAttention2,
     "sdpa": MiniCPMSdpaAttention,
 }