409 GB

Ctrl+K

1 contributor

History: 4 commits

AH22-neb

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values

f908ee8 verified 4 months ago

encoding
Upload folder using huggingface_hub 4 months ago
.gitattributes
1.58 kB
Upload folder using huggingface_hub 4 months ago
README.md
31 Bytes
initial commit 4 months ago
config.json
2.62 kB
Fix ignore pattern: .mlp.gate -> *.mlp.gate to avoid excluding gate_up_proj 4 months ago
conversion_report.json
1.77 kB
Upload folder using huggingface_hub 4 months ago
generation_config.json
171 Bytes
Upload folder using huggingface_hub 4 months ago
model-00001-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00002-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00003-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00004-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00005-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00006-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00007-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00008-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00009-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00010-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00011-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00012-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00013-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00014-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00015-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00016-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00017-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00018-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00019-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00020-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00021-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00022-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00023-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00024-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00025-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00026-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00027-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00028-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00029-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00030-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00031-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00032-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00033-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00034-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00035-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00036-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00037-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00038-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00039-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00040-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00041-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00042-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00043-of-000163.safetensors
5.37 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago
model-00044-of-000163.safetensors
5.35 GB
xet

Fix: dequantize attention weights (kv_b_proj, q_b_proj, etc.) - were stored as raw FP8 values 4 months ago