File size: 2,039 Bytes
b3a65ec | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 | {
"key": "5bit",
"mode": "affine",
"bits": 5,
"group_size": 64,
"scope": "Complete artifact and native dequantized-value checks, plus eight authored cached greedy fixtures; not a broad benchmark.",
"backend": "Linux native MLX CUDA 12 on NVIDIA B200",
"mac_acceptance": "pending",
"weight_bytes": 23838823629,
"files": [
{
"file": "model-00001-of-00005.safetensors",
"bytes": 5362048467,
"sha256": "88e2aced3d2273877135f295a2d11cbd48049956613dad4065fb7bea1eb5db00"
},
{
"file": "model-00002-of-00005.safetensors",
"bytes": 5215907743,
"sha256": "42727f19852b6f7cf33f3dc714a09f7b9c85d9a90fa32fe7d9937e092dfab7da"
},
{
"file": "model-00003-of-00005.safetensors",
"bytes": 5207519082,
"sha256": "0a5b21f2449651d4079aaf747651bdbacfb3b68bbff744e029bca997a8cde77d"
},
{
"file": "model-00004-of-00005.safetensors",
"bytes": 5203037976,
"sha256": "4010416518fbd293a7744289f8e8edeba8f0c0b9170933caee3d8ff768ba30cb"
},
{
"file": "model-00005-of-00005.safetensors",
"bytes": 2850310361,
"sha256": "9442f609345540abbe0f0ecfa541fe19d42c603598dfa775ee399ed63c4ede1d"
}
],
"strict_stock_reload": true,
"all_stored_floating_values_finite": true,
"all_dequantized_values_finite": true,
"all_generation_logits_finite": true,
"functional_passed": 8,
"functional_total": 8,
"tokenizer_source_files": {
"tokenizer.json": "06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523",
"tokenizer_config.json": "792fa3f0cb88b111e54ef3134c873531008c4df471d108da17903426e308aa7b",
"chat_template.jinja": "e84f32a23fdda27689f868aa4a1a5621f41133e51a48d7f3efcbea2839574259"
},
"quantized_modules": 512,
"dequantized_rows_checked": 32631720,
"http_smoke": {
"passed": 2,
"total": 2,
"endpoint": "/v1/chat/completions"
},
"held_out_perplexity": 8.307455349732711,
"paired_bf16_perplexity": 8.307409052128223,
"held_out_ppl_ratio_to_bf16": 1.0000055730498159
}
|