GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
neuralmagic
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
neuralmagic
Model Tree
deepseek-ai/DeepSeek-R1-Distill-Llama-8B
Input Modalities
Output Modalities
Supported Functionality
Quantized version of DeepSeek-R1-Distill-Llama-8B.
This model was obtained by quantizing the weights and activations of DeepSeek-R1-Distill-Llama-70B to FP8 data type. This optimization reduces the number of bits per parameter from 16 to 8, reducing the disk size and GPU memory requirements by approximately 50%.
Only the weights and activations of the linear operators within transformers blocks are quantized. Weights are quantized using a symmetric per-channel scheme, whereas quantizations are quantized using a symmetric per-token scheme. LLM Compressor is used for quantization.
This model can be deployed efficiently using the vLLM backend, as shown in the example below.
python
from transformers import AutoTokenizerfrom vllm import LLM, SamplingParamsnumber_gpus = 1model_name = "neuralmagic/DeepSeek-R1-Distill-Llama-8B-FP8-dynamic"tokenizer = AutoTokenizer.from_pretrained(model_name)sampling_params = SamplingParams(temperature=0.6, max_tokens=256, stop_token_ids=[tokenizer.eos_token_id])llm = LLM(model=model_name, tensor_parallel_size=number_gpus, trust_remote_code=True)messages_list = [[{"role": "user", "content": "Who are you? Please respond in pirate speak!"}],]prompt_token_ids = [tokenizer.apply_chat_template(messages, add_generation_prompt=True) for messages in messages_list]outputs = llm.generate(prompt_token_ids=prompt_token_ids, sampling_params=sampling_params)generated_text = [output.outputs[0].text for output in outputs]print(generated_text)
vLLM also supports OpenAI-compatible serving. See the documentation for more details.
This model was created with llm-compressor by running the code snippet below.
python
from transformers import AutoModelForCausalLM, AutoTokenizerfrom llmcompressor.modifiers.quantization import QuantizationModifierfrom llmcompressor.transformers import oneshotimport os# Load modelmodel_stub = "deepseek-ai/DeepSeek-R1-Distill-Llama-8B"model_name = model_stub.split("/")[-1]model = AutoModelForCausalLM.from_pretrained(model_stub,torch_dtype="auto",)tokenizer = AutoTokenizer.from_pretrained(model_stub)# Configure the quantization algorithm and schemerecipe = QuantizationModifier(targets="Linear",scheme="FP8_DYNAMIC",ignore=["lm_head"],)# Apply quantizationoneshot(model=model,recipe=recipe,)# Save to disk in compressed-tensors formatsave_path = model_name + "-FP8-dynamicmodel.save_pretrained(save_path)tokenizer.save_pretrained(save_path)print(f"Model and tokenizer saved to: {save_path}")
The model was evaluated on OpenLLM Leaderboard V1 and V2, using the following commands:
OpenLLM Leaderboard V1:
markdown
lm_eval \--model vllm \--model_args pretrained="neuralmagic/DeepSeek-R1-Distill-Llama-8B-FP8-dynamic",dtype=auto,max_model_len=4096,enable_chunked_prefill=True \--tasks openllm \--write_out \--batch_size auto \--output_path output_dir \--show_config
OpenLLM Leaderboard V2:
markdown
lm_eval \--model vllm \--model_args pretrained="neuralmagic/DeepSeek-R1-Distill-Llama-8B-FP8-dynamic",dtype=auto,max_model_len=4096,enable_chunked_prefill=True \--apply_chat_template \--fewshot_as_multiturn \--tasks leaderboard \--write_out \--batch_size auto \--output_path output_dir \--show_config
This model achieves up to 1.4x speedup in single-stream deployment and up to 1.3x speedup in multi-stream asynchronous deployment, depending on hardware and use-case scenario. The following performance benchmarks were conducted with vLLM version 0.7.2, and GuideLLM.
markdown
guidellm --model neuralmagic/DeepSeek-R1-Distill-Llama-8B-FP8-dynamic --target "http://localhost:8000/v1" --data-type emulated --data "prompt_tokens=<prompt_tokens>,generated_tokens=<generated_tokens>" --max seconds 360 --backend aiohttp_server
**Use case profiles: prompt tokens / generation tokens
**QPD: Queries per dollar, based on on-demand cost at Lambda Labs (observed on 2/18/2025).
**Use case profiles: prompt tokens / generation tokens
**QPS: Queries per second.
**QPD: Queries per dollar, based on on-demand cost at Lambda Labs (observed on 2/18/2025).