Instructions to use PanocularAI/PanoLM-380M-Base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use PanocularAI/PanoLM-380M-Base with Transformers:

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

pipe = pipeline("text-generation", model="PanocularAI/PanoLM-380M-Base", trust_remote_code=True)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("PanocularAI/PanoLM-380M-Base", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use PanocularAI/PanoLM-380M-Base with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "PanocularAI/PanoLM-380M-Base"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "PanocularAI/PanoLM-380M-Base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/PanocularAI/PanoLM-380M-Base

SGLang

How to use PanocularAI/PanoLM-380M-Base 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 "PanocularAI/PanoLM-380M-Base" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "PanocularAI/PanoLM-380M-Base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "PanocularAI/PanoLM-380M-Base" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "PanocularAI/PanoLM-380M-Base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use PanocularAI/PanoLM-380M-Base with Docker Model Runner:
```
docker model run hf.co/PanocularAI/PanoLM-380M-Base
```

sd24 commited on May 31

Commit

94c2e8a

verified ·

1 Parent(s): a484c90

Upload folder using huggingface_hub

Browse files

Files changed (3) hide show

README.md +17 -12
chat_template.jinja +5 -0
sharded/shard-00001-model-00001-of-00001.safetensors +1 -1

README.md CHANGED Viewed

@@ -6,13 +6,17 @@ language:
 pipeline_tag: text-generation
 tags:
   - panolm
-  - pda
   - linear-attention
 ---
 ## Training Data
-Pretrained on 300B tokens, a weighted mixture of three web-scale English corpora:
 | Weight | Dataset                |
 |-------:|------------------------|
@@ -30,11 +34,14 @@ flash-linear-attention==0.5.0
 ## Usage
 ```python
 from transformers import AutoModelForCausalLM, AutoTokenizer
 model = AutoModelForCausalLM.from_pretrained(
-    "PanocularAI/PanoLM-380M",
     trust_remote_code=True,
 ).cuda()  # fla's RMSNorm uses Triton kernels that only run on CUDA tensors.
 print(model)
@@ -47,9 +54,7 @@ inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
 outputs = model.generate(
     **inputs,
     max_length=512,
-    temperature=1.0,
-    top_k=50,
-    top_p=1.0,
     use_cache=False,
     do_sample=True,
     trust_remote_code=True,
@@ -79,9 +84,9 @@ metric the suite was designed around.
 | Task           | Metric   |  Value |
 |----------------|----------|-------:|
-| drop     | contains | 0.2286 |
-| fda      | contains | 0.0499 |
-| nq_2048  | contains | 0.0687 |
-| squadv2  | contains | 0.3542 |
-| swde     | contains | 0.2439 |
-| triviaqa | contains | 0.4680 |

 pipeline_tag: text-generation
 tags:
   - panolm
+  - kda
   - linear-attention
 ---
+# PanoLM-380M
+This repo contains the base 370M PanoLM model, which is a linear-attention causal language.
 ## Training Data
+Pretrained on a weighted mixture of three web-scale English corpora:
 | Weight | Dataset                |
 |-------:|------------------------|
 ## Usage
+Replace `<repo_id>` with the HF Hub identifier (e.g. `your-org/PanoLM-370M`).
 ```python
 from transformers import AutoModelForCausalLM, AutoTokenizer
+repo_id = "<repo_id>"
 model = AutoModelForCausalLM.from_pretrained(
+    repo_id,
     trust_remote_code=True,
 ).cuda()  # fla's RMSNorm uses Triton kernels that only run on CUDA tensors.
 print(model)
 outputs = model.generate(
     **inputs,
     max_length=512,
+    top_k=10,
     use_cache=False,
     do_sample=True,
     trust_remote_code=True,
 | Task           | Metric   |  Value |
 |----------------|----------|-------:|
+| based_drop     | contains | 0.2286 |
+| based_fda      | contains | 0.0499 |
+| based_nq_2048  | contains | 0.0687 |
+| based_squadv2  | contains | 0.3542 |
+| based_swde     | contains | 0.2439 |
+| based_triviaqa | contains | 0.4680 |

chat_template.jinja ADDED Viewed

	@@ -0,0 +1,5 @@

+{% if messages[0]['role'] != 'system' %}{{ 'A chat between a curious user and an artificial intelligence assistant. The
+assistant gives helpful, detailed, and polite answers to the user\'s questions. ' }}{% endif %}{% for message in
+messages %}{% if message['role'] == 'system' %}{{ message['content'] + ' ' }}{% elif message['role'] == 'user' %}{{
+'USER: ' + message['content'] + ' ' }}{% elif message['role'] == 'assistant' %}{{ 'ASSISTANT: ' + message['content'] +
+'</s>' }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ 'ASSISTANT: ' }}{% endif %}

sharded/shard-00001-model-00001-of-00001.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ab16d8ee6e184fc2adf598c1cee1a7c424a2fe4237b6d71572128a320a214b9b
 size 1549115704

 version https://git-lfs.github.com/spec/v1
+oid sha256:bd0946e12a7ea0d51a3d1f827d07325d1e3e3fbca73b54d73398eb5955d7983d
 size 1549115704