Insop
Insop
- testing tested with chroma==0.4.17 *Note*: this change is not related to the new cohort. Thank you and let me know your feedback, @okhat
Update together model prompt and stop sequence - update prompt based on the model name - update stop sequence *TEST*: - the following models are tested using `intro.ipynb` - completion...
[WIP] Will update some more markdown files before removing the `[WIP]` This PR adds descriptions for Ollama and Together model usages in the notebook and md file. Hi @okhat PTAL...
Since https://github.com/stanfordnlp/dspy/issues/521 is closed state, creating a new issue. Now, the initial PR https://github.com/stanfordnlp/dspy/pull/522 for the issue was reverted by this PR, https://github.com/stanfordnlp/dspy/pull/584 recently, I think we need to keep...
Hi, thank you for great work (paper and repo)! Even with the dspy commit hash, I am not able to locate `Models` from dspy repo. Let me know how I...
The current `check_two_equal` may not able to check the `NaN` case. We can improve it by using `std::isnan`. When the following incorrect SIMD implementation runs on M1 environment, the test...
This branch includes a micro-optimization for `softmax_forward_kernel5`. ## Summary - ~~use `warpReduceMax` in `attention_forward.cu` to use `__shfl_down_sync` to be consistent with the other kernels (reduce to all threads in a...
### Changes Add `repkv_backward_kernel2` - improve `repkv_backward_kernel1 `by reducing thread used per @karpathy's suggestion Also add `repkv_kernel2` simiar to `backward_kernel2` ### Here is the test output for `repkv_backward_kernel2` ``` #...
### Is there an existing issue for the same bug? - [x] I have checked the existing issues. ### Describe the bug and reproduction steps Followed a quick start guide...
#### Context What is the purpose of this PR? Is it to - [ ] add a new feature - [ ] fix a bug - [ ] update tests...