Glaciohound/LM-Infinite
Implementation of paper "LM-Infinite: Simple On-the-Fly Length Generalization for Large Language Models"
PythonMIT
Issues
- 0
passkey代码运行不通
#10 opened by xjwhy - 0
kv_seq_len bug?
#9 opened by chenlidar - 0
- 4
TypeError: attn_forward_factory() missing 5 required positional arguments: 'top_k_attention', 'top_k_insert_at', 'top_k_from_layer', 'top_k_to_layer', and 'layer_i'
#7 opened by canelxie - 1
Some errors.
#6 opened by chaochen99 - 1
- 1
Should the llama model be fine-tuned?
#5 opened by yinwangsong - 2
Implementation with RoPE
#4 opened by sdc17 - 1
How to Inferance?
#2 opened by farrael004 - 2
GPTNeoX or Transformers support?
#1 opened by fblgit