runanywhere
/

llama3_2_1b_HNPU

Text Generation

Model card Files Files and versions

llama3_2_1b_HNPU / v79

Ctrl+K

Ctrl+K

3 contributors

History: 20 commits

Aman0runanywhere's picture

Aman0runanywhere

revert to working P=128 while a16 int16-calib is tuned

6caa332 verified 5 days ago

llama-3.2-1b.json

1.34 kB
revert manifest to llama_full_wqo prefill (drop-in fp16 test uses same graph name) 5 days ago
llama_dec_wqo_o3.bin
Pickle imports
- No problematic imports detected
What is a pickle import?
981 MB
xet

Upload v79/llama_dec_wqo_o3.bin with huggingface_hub 27 days ago
llama_embed_f16.bin
Pickle imports
- No problematic imports detected
What is a pickle import?
525 MB
xet

Upload v79/llama_embed_f16.bin with huggingface_hub 27 days ago
llama_full_f16.bin
Pickle imports
- No problematic imports detected
What is a pickle import?
1.96 GB
xet

fp16-P=512 prefill DIAGNOSTIC bin (user-authorized W8-vs-drift test; revert after) 5 days ago
llama_full_wqo_o3.bin
Pickle imports
- No problematic imports detected
What is a pickle import?
982 MB
xet

revert to working P=128 while a16 int16-calib is tuned 5 days ago
llama_lmhead.bin
Pickle imports
- No problematic imports detected
What is a pickle import?
526 MB
xet

Upload v79/llama_lmhead.bin with huggingface_hub 27 days ago
tokenizer.json

9.09 MB
Upload v79/tokenizer.json with huggingface_hub 27 days ago