 Tobi Lutke
|
891f3262cf
Fix GRPO reward function to handle think blocks and end tokens
|
4 kuukautta sitten |
|
Tobi Lutke
|
66bb8ed963
Remove beads reference from CLAUDE.md
|
4 kuukautta sitten |
|
Tobi Lutke
|
2267986302
Remove beads issue tracking
|
4 kuukautta sitten |
|
Tobi Lutke
|
8a1c4cdab0
Add 1.7B and 4B GRPO training and GGUF conversion scripts
|
4 kuukautta sitten |
|
Tobi Lutke
|
b9b1b39a76
Update README with separate model repos
|
4 kuukautta sitten |
|
Tobi Lutke
|
312c281109
Update README for unified model repository structure
|
4 kuukautta sitten |
|
Tobi Lutke
|
2648512b7c
Fix TUI to load GRPO models with SFT base first
|
4 kuukautta sitten |
|
Tobi Lutke
|
f96766cce8
Fix GRPO model loading to use SFT base first
|
4 kuukautta sitten |
|
Tobi Lutke
|
f6a6716c44
Refactor evals into separate run and score scripts
|
4 kuukautta sitten |
|
Tobi Lutke
|
857a85ab58
Clean up evaluation files
|
4 kuukautta sitten |
|
Tobi Lutke
|
dc8f5a2335
Strict format validation: every line must be lex:/vec:/hyde:
|
4 kuukautta sitten |
|
Tobi Lutke
|
2ad507a86e
Add chat template leakage detection to reward function
|
4 kuukautta sitten |
|
Tobi Lutke
|
6062dc769f
Add named entity extraction to GRPO reward function
|
4 kuukautta sitten |
|
Tobi Lutke
|
32706a720f
Refactor finetune folder: train/rl scripts with YAML configs
|
4 kuukautta sitten |
|
Tobi Lutke
|
d32e13c172
Add HuggingFace login and comprehensive scoring to GRPO v2 training
|
4 kuukautta sitten |
|
Tobi Lutke
|
c35dbd6cbd
Add comprehensive scoring system for query expansion
|
4 kuukautta sitten |
|
Tobi Lutke
|
994a094546
Update README with final evaluation results
|
4 kuukautta sitten |
|
Tobi Lutke
|
0353994e7d
Fix GRPO training script for TRL API compatibility
|
4 kuukautta sitten |
|
Tobi Lutke
|
7cca164dd9
Add query expansion model finetuning infrastructure
|
4 kuukautta sitten |
|
komsit37
|
88f78314bb
Fix sqlite-vec loading with BREW_PREFIX (#42)
|
4 kuukautta sitten |
|
Tobias Lütke
|
3c7dfad1b6
Make docid lookup more lenient with quotes support (#39)
|
4 kuukautta sitten |
|
Joshua Lelon Mitchell
|
fbd7fe8c8e
Fix docid lookup in qmd get command (#36)
|
4 kuukautta sitten |
|
Tobias Lütke
|
5b1671d2f6
Merge pull request #38 from odysseus0/fix/readme-model-sizes
|
4 kuukautta sitten |
|
George Zhang
|
c8f72de12e
docs: fix query expansion model size (Qwen3-1.7B, not 0.6B)
|
4 kuukautta sitten |
|
Tobi Lutke
|
7817dc11a4
Show embedding notice only once at end of qmd update
|
4 kuukautta sitten |
|
Tobias Lütke
|
6fbad4e9a6
Merge pull request #15 from gavrix/main
|
4 kuukautta sitten |
|
Tobias Lütke
|
77e1f82cd9
Merge pull request #23 from mbrendan/fix-vsearch-hang
|
4 kuukautta sitten |
|
Tobias Lütke
|
2bde6804fb
Merge pull request #29 from mostlydev/fix-mcp-server-exit
|
4 kuukautta sitten |
|
Wojtek
|
a0ca60bc6f
Fix MCP server exiting immediately after startup
|
4 kuukautta sitten |
|
Brendan McCord
|
aea494bb24
Fix regression test to use proper test helpers
|
4 kuukautta sitten |
