Giters
Cornell-RelaxML
/
quip-sharp
Geek Repo:
Geek Repo
Github PK Tool:
Github PK Tool
Stargazers:
439
Watchers:
10
Issues:
44
Forks:
33
Cornell-RelaxML/quip-sharp Issues
model size confirmation
Closed
3 months ago
Comments count
3
Procedures for quantizing generic architectures
Closed
3 months ago
Comments count
1
Group-wise Quantization
Closed
3 months ago
Comments count
1
Pytorch dequantization
Closed
3 months ago
Comments count
3
NameError: name 'quant_emb' is not defined
Closed
4 months ago
Comments count
5
How many samples do you use in checkpoints?
Closed
4 months ago
Comments count
2
Llama-2-7b-E8P-2Bit not loading correctly.
Closed
4 months ago
Comments count
1
HF Mistral-7B and Llama 2 7b chat Not working.
Closed
5 months ago
Comments count
10
TypeError: decompress_e8p_origorder(): incompatible function arguments.
Closed
5 months ago
Comments count
10
3 bit quantization
Closed
6 months ago
Comments count
4
There are some issues when I try to run the Yi34b model with 2bits quant
Closed
5 months ago
Comments count
4
What could went wrong
Closed
6 months ago
Comments count
3
Exception: Saved weights version (0) does not match the codebook version (1).
Closed
5 months ago
Comments count
6
support qwen model
Closed
5 months ago
Comments count
38
llamafied model have some issues happening in hfize_llama.py
Closed
5 months ago
Comments count
22
ROCm Build Error
Updated
5 months ago
Comments count
3
Question about error proxy in show_metrics
Closed
6 months ago
Comments count
6
In the same vein as #17
Closed
6 months ago
Comments count
1
trouble building quip_tools
Closed
6 months ago
Comments count
3
Problem with namespace nvcuda
Closed
6 months ago
Comments count
6
distribute the memory usage evenly across both cards?
Closed
6 months ago
Comments count
2
Does QUIP# support deepseek-llm-chat-67b?
Closed
6 months ago
Comments count
6
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions. Same issue as #4
Closed
6 months ago
Comments count
4
FileNotFoundError: [Errno 2] No such file or directory: '/worker/quip_llama2/hessians/2_qkv.pt'
Closed
6 months ago
Comments count
2
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
Closed
6 months ago
Comments count
4
mistral-7b
Closed
6 months ago
Comments count
7
mixtral 8x7b
Closed
6 months ago
Comments count
1
why don't you do the register_buffer inside the QuantizedLinear() init ?
Closed
6 months ago
Comments count
3
Load LORA?
Closed
6 months ago
Comments count
1
custom 1.3B llama quant
Closed
6 months ago
Comments count
22
How to quant 1.3B model to 2bit
Closed
6 months ago
Comments count
3
Cannot re-initialize CUDA in forked subprocess
Closed
6 months ago
Comments count
2
Low Ppl benchmark results
Closed
6 months ago
Comments count
1
[Request] Pre-Converted files for Yi-34B-200k
Closed
6 months ago
Comments count
1
Mistral models output gibberish
Closed
6 months ago
Comments count
1
can you provide example or code to compute hessians
Closed
6 months ago
Comments count
2