rewrite flm-add in cpp as flm add - #34
Closed
Atomic-Germ wants to merge 3 commits into
Closed
Atomic-Germ wants to merge 3 commits into
Atomic-Germ wants to merge 3 commits into
Conversation
Collaborator
|
We also have to think .. if we are supporting gguf, then hugging face standard formats and everything else become a possibility. I'm not sure what the future of this is, but good for now. |
Cyronius
self-requested a review
September 8, 2026 23:42
Cyronius
approved these changes
Sep 8, 2026
bool add_thinking = false; // --thinking -> details.think bool add_think_toggleable = false; // --think-toggleable -> details.think_toggleable std::string add_parameter_size = ""; // --parameter-size -> details.parameter_size std::string add_quantization = ""; // --quantization -> details.quantization_level int add_context_length = -1; // --context-length -> default_context_length int add_max_prefill = -1; // --max-prefill -> max_prefill_len std::string add_label = ""; // --label (comma-separated) -> label[] bool add_list_families = false; // --list-families -> list available family open_kernels
Owner
Author
@Cyronius - I've had that on my mind for a while now; I think that will be the single most powerful thing we can bring to this, and it'll be sooner rather than later I think. |
vegah
added a commit
to vegah/OpenFlowLM-Next
that referenced
this pull request
Sep 14, 2026
…odel_info.json Scope is now only which models the application knows and where they are. Kernel lookup, the Python oflm-add and q4nx-build changes, and the history path are reverted to main: Atomic-Germ#34 replaces oflm-add with `oflm add`, which links fine-tunes to their family's open kernels and needs no per-name xclbin lookup. Added what Atomic-Germ#34 needs to work without exports: registry_directories() includes OFLM_MODEL_PATH, where `oflm add` writes model_list.json and model_info.json, and find_model_infos() / model_registry::load_model_info() merge user model_info.json files over the shipped one by tag (a user file never replaces a shipped tag's hashes). The downloader reads the merged file at all three call sites. Tests: user_dirs_test, 64 checks on MSVC and 62 on g++ 13 / C++20; the three model_info rules were each seen to fail with their fix disabled. With no OFLM_CONFIG_PATH and OFLM_MODEL_PATH laid out the way `oflm add` writes it, `oflm check <tag>` verifies every file; without that model_info.json it fails with `key '<tag>' not found`. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Owner
Author
|
Superceded by #86 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This addresses #29 ; needs extensive testing on different systems -- not sure if flm pull should be outright replaced yet. This will need quite a bit done when gguf hits but that's alright.