Ecosyste.ms: Timeline
Browse the timeline of events for every public repo on GitHub. Data updated hourly from GH Archive.
ngxson pushed 1 commit to master ngxson/llama.cpp
- vulkan: multi-row k quants (#10846) * multi row k quant shaders! * better row selection * more row choices ... d79d8f3
ngxson pushed 1 commit to master ngxson/llama.cpp
- examples, ggml : fix GCC compiler warnings (#10983) Warning types fixed (observed under MSYS2 GCC 14.2.0): * format... d283d02
ngxson created a comment on an issue on ggerganov/llama.cpp
I think that the time it takes to generate internal json is too small compared to the time it takes to send a big json from server to client (network constraint), so to me it's OK to only filter be...
ngxson closed a pull request on ggerganov/llama.cpp
server: add OpenAI compatible response format for legacy /completions with b…
This is based of a previous [PR](https://github.com/ggerganov/llama.cpp/pull/10627) However, @ngxson seems to be working refactoring the server.cpp to prevent use of JSON as stated [here](https:...
ngxson opened a draft pull request on ggerganov/llama.cpp
server : add OAI compat for /v1/completions
Supersede #10645 Ref documentation: https://platform.openai.com/docs/api-reference/completions/object
ngxson pushed 1 commit to master ngxson/llama.cpp
- server : add support for "encoding_format": "base64" to the */embeddings endpoints (#10967) * add support for base64... 9ba399d
ngxson pushed 1 commit to master ggerganov/llama.cpp
- server : add support for "encoding_format": "base64" to the */embeddings endpoints (#10967) * add support for base64... 9ba399d
ngxson closed an issue on ggerganov/llama.cpp
Feature Request: support `"encoding_format": "base64"` in the `*/embeddings` endpoints
### Prerequisites - [X] I am running the latest code. Mention the version if possible as well. - [X] I carefully followed the [README.md](https://github.com/ggerganov/llama.cpp/blob/master/README....
ngxson closed a pull request on ggerganov/llama.cpp
server : add support for "encoding_format": "base64" to the */embeddings endpoints
# Add base64 encoding format support for embeddings endpoints ## Overview This PR implements support for base64 format in embedding responses, aligning with OpenAI's API functionality and impro...
ngxson pushed 1 commit to master ngxson/llama.cpp
- ggml : more perfo with llamafile tinyblas on x86_64 (#10714) * more perfo with llamafile tinyblas on x86_64. - ad... 2cd43f4
ngxson pushed 1 commit to master ngxson/llama.cpp
- server: allow filtering llama server response fields (#10940) * llama_server_response_fields * llama_server_resp... 09fe2e7
ngxson pushed 1 commit to master ggerganov/llama.cpp
- server: allow filtering llama server response fields (#10940) * llama_server_response_fields * llama_server_resp... 09fe2e7
ngxson closed a pull request on ggerganov/llama.cpp
server: allow filtering llama server response fields
This adds a new feature. Allow Filtering LLama Server Response Fields from #10819. Usage: Curl: ```bash curl --request POST \ --url http://localhost:8080/completion \ --header "Conte...
ngxson created a review comment on a pull request on ggerganov/llama.cpp
```py res = server.make_request("POST", "/completion", data={ "prompt": "I believe the meaning of life is", "seed": 42, "temperature": 1.0, # <==== Change to...
ngxson created a review comment on a pull request on ggerganov/llama.cpp
Does it work if here we set temperature to 0.0 instead of forcing it to 1 slot?
ngxson created a review comment on a pull request on ggerganov/llama.cpp
It's a bad practice to use try..catch inside a test, because if one of the `assert` fails, it won't let you know it happens at exactly which line of code.
ngxson created a review comment on a pull request on ggerganov/llama.cpp
The current behavior is ok, but just need to be more clear when writing docs. `return an empty json` means `{}`, it's not the current behavior. I rephrased this to be more clear.