summaryrefslogtreecommitdiff
path: root/models
AgeCommit message (Expand)Author
2024-09-30<debug>Ashwin Bharambe
2024-09-30Update MODEL_CARD - Feedback links/format (#153)machina-source
2024-09-30Minor typos (#154)Mark Sze
2024-09-25small fix on the model card (#151)Kai Wu
2024-09-25Create eval_details.mdRohit Patel
2024-09-25Support for Llama3.2 series of models (#150)Ashwin Bharambe
2024-09-24Update MODEL_CARD.mdJoseph Spisak
2024-09-23[API updates] (#146)Ashwin Bharambe
2024-09-15chore: update schema_utils.py (#142)Ikko Eltociear Ashimine
2024-09-14Remove duplicated target computation in generation.py (#139)Yufeng Zhang
2024-09-14Add an explicit `pth_file_count` fieldAshwin Bharambe
2024-09-13update vocab size for llama2 modelsAshwin Bharambe
2024-09-13Drop confusing hardware_requirements from SKUAshwin Bharambe
2024-09-04Only strip out assistant headerAshwin Bharambe
2024-09-04Make sure prompt does not have trailing whitespaceAshwin Bharambe
2024-09-04Fix importAshwin Bharambe
2024-09-03API updates (#132)Ashwin Bharambe
2024-08-27remove pydantic pinning from another requirements.txtAshwin Bharambe
2024-08-24fix max_seq_len for llama_guard_2Ashwin Bharambe
2024-08-22Another fix for llama-guard-2Ashwin Bharambe
2024-08-22Llama2 download paths are lower case :/Ashwin Bharambe
2024-08-21Fix docstring args names (#119)Sergii Dymchenko
2024-08-21Added Llama Guard 2 8b sku (#117)varunfb
2024-08-20Use tokenizer.model from the repository for the example scriptsAshwin Bharambe
2024-08-19Move code from llama3_1 -> llama3Ashwin Bharambe
2024-08-15Add `WebMethod.method` for overriding the HTTP method for a routeAshwin Bharambe
2024-08-15updating tool call formats in llama cli (#110)Hardik Shah
2024-08-15note inclusion of prev Q&A pairs in prompt for SQuAD eval in eval_details.md ...Jason Krone
2024-08-14update cli system prompt templatesHardik Shah
2024-08-14Add example scripts to show how to run the model (#108)Ashwin Bharambe
2024-08-13added micro avg metrics for MMLU-Pro to eval_details.md (#106)Kai Wu
2024-08-13update system prompt to drop newlineHardik Shah
2024-08-12upgrade pydantic to latestHardik Shah
2024-08-08Fix n_kv_heads for mp16 modelsAshwin Bharambe
2024-08-08Merge distros to main (#97)Dalton Flanagan
2024-08-07Improve model list display (#94)Dalton Flanagan
2024-08-07SKUs: chat models categorized as instruct modelsdltn
2024-08-07Add Llama 2 and Llama 3 SKUsdltn
2024-08-07Add checklist.chk to files which are downloadedAshwin Bharambe
2024-08-07parse json for tool calls properlyHardik Shah
2024-08-06mmlu_pro metric should be macro_avg/acc instead of micro_avg/acc_char (#92)Kai Wu
2024-08-05test for custom tool callingHardik Shah
2024-07-31Change how we name SKUs; add safety SKUs (#79)Ashwin Bharambe
2024-07-31Add missing fp8_scales files to `meta-llama-3.1-405b-fp8` (#69)pmysl
2024-07-31fix: chmod +x download.sh (#72)Matthew Berryman
2024-07-31Remove md5sum requirement (#74)minimalic
2024-07-30move model SKU stuff upwards into top-level llama_models namespaceAshwin Bharambe
2024-07-30Hard code consolidated.XX.pth file sizes for 405B due to CloudfrontAshwin Bharambe
2024-07-29add recommended sampling paramsAshwin Bharambe
2024-07-29add a mapping from SKU -> llamameta.net subfoldersAshwin Bharambe
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2024-10-22 22:17:45 +0000