ConstDemi commited on
Commit
df80fd1
·
verified ·
1 Parent(s): d37921d

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +0 -2
README.md CHANGED
@@ -19,8 +19,6 @@ pipeline_tag: text-generation
19
 
20
  4-bit AWQ quantization of [Vikhrmodels/Vikhr-Nemo-12B-Instruct-R-21-09-24](https://huggingface.co/Vikhrmodels/Vikhr-Nemo-12B-Instruct-R-21-09-24), a Russian-language grounded-RAG model. Produced with [llm-compressor](https://github.com/vllm-project/llm-compressor) in the compressed-tensors format for vLLM.
21
 
22
- Main motivation: on pre-Ampere GPUs (sm75, e.g. Quadro RTX 6000 / RTX 20xx) GPTQ checkpoints fall back to slow non-Marlin kernels in vLLM, while this checkpoint runs via `MarlinLinearKernel for CompressedTensorsWNA16`, giving a large speedup in prefill and batched decode.
23
-
24
  ## Quantization details
25
 
26
  - Method: AWQ, scheme W4A16, `lm_head` kept unquantized
 
19
 
20
  4-bit AWQ quantization of [Vikhrmodels/Vikhr-Nemo-12B-Instruct-R-21-09-24](https://huggingface.co/Vikhrmodels/Vikhr-Nemo-12B-Instruct-R-21-09-24), a Russian-language grounded-RAG model. Produced with [llm-compressor](https://github.com/vllm-project/llm-compressor) in the compressed-tensors format for vLLM.
21
 
 
 
22
  ## Quantization details
23
 
24
  - Method: AWQ, scheme W4A16, `lm_head` kept unquantized