bielik-q2-sharp / variant-e /quantization_config.txt
Jakubrd4's picture
Upload variant-e full results (quant logs + eval 5-shot/0-shot MC/GEN)
acc64cf verified
=== VPTQ Quantization Configuration ===
Date: 2026-02-22 19:40:24 UTC
Target: ~2-bit quantization with Residual VQ
Model: speakleash/Bielik-11B-v2.3-Instruct (MistralForCausalLM)
Parameters:
--model_name /workspace/models/bielik-11b-instruct
--output_dir /workspace/variant-e/output/
--vector_lens -1 8 # Skip embeddings (-1), quantize in vectors of 8
--group_num 1 # Single group
--num_centroids -1 65536 # 2^16 centroids = 16 bits per vector index
--num_res_centroids -1 256 # 2^8 residual centroids
--npercent 0 # No outlier channels
--blocksize 128 # Block size for quantization
--new_eval # Use new evaluation method
--seq_len 4096 # Sequence length for calibration
--kmeans_mode hessian # Hessian-weighted k-means clustering
--num_gpus 1 # Single GPU
--enable_perm # Channel permutation for better quant
--enable_norm # Channel normalization
--save_model # Save quantized model
--save_packed_model # Save packed model for inference
--hessian_path /workspace/hessians/quip-format/hessians
--kiter 100 # K-means iterations
--ktol 1e-5 # K-means convergence tolerance
Effective bitwidth calculation:
Primary: log2(65536) / 8 = 16/8 = 2.0 bits/weight
Residual: log2(256) / 8 = 8/8 = 1.0 bits/weight overhead
Total: ~2.0 bits/weight (residual adds codebook storage overhead, not per-weight)
Hessian source: Jakubrd4/bielik-quip-e8p12
Format: QuIP# (flatH + mu + n + ct per layer)
Calibration: CulturaX-PL, 512 samples x 4096 tokens
Layers: 50 x 4 projections (qkv, o, up, down) = 200 files