github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 00:22:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #6341] Llama 3.1 70B high-quality HQQ quantized model - 99%+ quality of fp16 #3980

New Issue
Open
opened 2026-04-12 14:51:07 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-12 14:51:07 -05:00
Owner
Copy Link

Originally created by @gileneusz on GitHub (Aug 13, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/6341

I'm not really sure if that's possible but adding that to ollama could really impact the performance on 4-bit quant option:

99%+ in all benchmarks in lm-eval relative performance to FP16 and similar inference speed to fp16

url:
https://huggingface.co/mobiuslabsgmbh/Llama-3.1-70b-instruct_4bitgs64_hqq

Screenshot 2024-08-13 at 19 03 57

also this:

https://huggingface.co/ModelCloud/Meta-Llama-3.1-70B-Instruct-gptq-4bit

Screenshot 2024-08-13 at 19 07 18
Originally created by @gileneusz on GitHub (Aug 13, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/6341 I'm not really sure if that's possible but adding that to ollama could really impact the performance on 4-bit quant option: 99%+ in all benchmarks in lm-eval relative performance to FP16 and similar inference speed to fp16 url: https://huggingface.co/mobiuslabsgmbh/Llama-3.1-70b-instruct_4bitgs64_hqq <img width="604" alt="Screenshot 2024-08-13 at 19 03 57" src="https://github.com/user-attachments/assets/64cd0427-b7c7-4fb8-b846-15f172669248"> also this: https://huggingface.co/ModelCloud/Meta-Llama-3.1-70B-Instruct-gptq-4bit <img width="597" alt="Screenshot 2024-08-13 at 19 07 18" src="https://github.com/user-attachments/assets/c3518ffe-323d-42f0-9162-d188179797fb">
GiteaMirror added the model label 2026-04-12 14:51:07 -05:00
GiteaMirror commented 2026-04-12 14:51:08 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Aug 13, 2024):

Requires support in llama.cpp. There have been issues raised (https://github.com/ggerganov/llama.cpp/issues/6368, https://github.com/ggerganov/llama.cpp/issues/4782) but unfortunately no progress.

<!-- gh-comment-id:2286873852 --> @rick-github commented on GitHub (Aug 13, 2024): Requires support in llama.cpp. There have been issues raised (https://github.com/ggerganov/llama.cpp/issues/6368, https://github.com/ggerganov/llama.cpp/issues/4782) but unfortunately no progress.
GiteaMirror commented 2026-04-12 14:51:08 -05:00
Author
Owner
Copy Link

@charlesrwest commented on GitHub (Aug 21, 2024):

I would also be really interested in seeing support added for this. Would pay $100 bounty, if that helps.

<!-- gh-comment-id:2301914632 --> @charlesrwest commented on GitHub (Aug 21, 2024): I would also be really interested in seeing support added for this. Would pay $100 bounty, if that helps.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label model
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#3980
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?