metadata

base_model: speakleash/Bielik-11B-v2.3-Instruct
license: apache-2.0
model_creator: speakleash
model_name: Bielik-11B-v2.3-Instruct
quantized_by: Second State Inc.
language:
  - pl
library_name: transformers
inference:
  parameters:
    temperature: 0.2

Bielik-11B-v2.3-Instruct-GGUF

Original Model

speakleash/Bielik-11B-v2.3-Instruct

Run with LlamaEdge

LlamaEdge version: v0.18.3

Prompt template

Prompt type: chatml

Prompt string

<|im_start|>system
{system_message}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

Context size: 32000

Run as LlamaEdge service

wasmedge --dir .:. --nn-preload default:GGML:AUTO:Bielik-11B-v2.3-Instruct-Q5_K_M.gguf \
  llama-api-server.wasm \
  --model-name Bielik-11B-v2.3-Instruct \
  --prompt-template chatml \
  --ctx-size 32000

Run as LlamaEdge command app

wasmedge --dir .:. --nn-preload default:GGML:AUTO:Bielik-11B-v2.3-Instruct-Q5_K_M.gguf \
  llama-chat.wasm \
  --prompt-template chatml \
  --ctx-size 32000

Quantized GGUF Models

Name	Quant method	Bits	Size	Use case
Bielik-11B-v2.3-Instruct-Q2_K.gguf	Q2_K	2	4.16 GB	smallest, significant quality loss - not recommended for most purposes
Bielik-11B-v2.3-Instruct-Q3_K_L.gguf	Q3_K_L	3	5.88 GB	small, substantial quality loss
Bielik-11B-v2.3-Instruct-Q3_K_M.gguf	Q3_K_M	3	5.40 GB	very small, high quality loss
Bielik-11B-v2.3-Instruct-Q3_K_S.gguf	Q3_K_S	3	4.85 GB	very small, high quality loss
Bielik-11B-v2.3-Instruct-Q4_0.gguf	Q4_0	4	6.32 GB	legacy; small, very high quality loss - prefer using Q3_K_M
Bielik-11B-v2.3-Instruct-Q4_K_M.gguf	Q4_K_M	4	6.72 GB	medium, balanced quality - recommended
Bielik-11B-v2.3-Instruct-Q4_K_S.gguf	Q4_K_S	4	6.36 GB	small, greater quality loss
Bielik-11B-v2.3-Instruct-Q5_0.gguf	Q5_0	5	7.70 GB	legacy; medium, balanced quality - prefer using Q4_K_M
Bielik-11B-v2.3-Instruct-Q5_K_M.gguf	Q5_K_M	5	7.91 GB	large, very low quality loss - recommended
Bielik-11B-v2.3-Instruct-Q5_K_S.gguf	Q5_K_S	5	7.70 GB	large, low quality loss - recommended
Bielik-11B-v2.3-Instruct-Q6_K.gguf	Q6_K	6	9.16 GB	very large, extremely low quality loss
Bielik-11B-v2.3-Instruct-Q8_0.gguf	Q8_0	8	11.9 GB	very large, extremely low quality loss - not recommended
Bielik-11B-v2.3-Instruct-f16.gguf	f16	16	22.3 GB

Quantized with llama.cpp b5201