 Tobias Lütke
|
55c951b15e
Merge pull request #349 from byheaven/fix/qwen3-embedding-model-filename-case
|
2 mesi fa |
|
Tobi Lutke
|
55f16460d0
fix(ci): guard LLM calls in CI and increase test timeouts
|
2 mesi fa |
|
Tobi Lutke
|
c68904fe08
refactor: move CLI and MCP to subdirectories, MCP consumes SDK
|
2 mesi fa |
|
YuBai
|
740b17b485
docs: fix Qwen3-Embedding GGUF filename case in README and llm.ts
|
2 mesi fa |
|
Tobi Lutke
|
ad38c1f698
feat: add intent parameter for query disambiguation
|
2 mesi fa |
|
Tobi Lutke
|
e3549dab1a
perf(rerank): cap parallelism, deduplicate chunks, cache by content
|
2 mesi fa |
|
Tobias Lütke
|
7904ab9a9d
Merge pull request #273 from daocoding/feature/configurable-embed-model
|
2 mesi fa |
|
Tobias Lütke
|
ee08997f23
Merge pull request #313 from 0xble/fix/expand-context-size-config
|
2 mesi fa |
|
Brian Le
|
0dec1df047
fix(llm): make expansion context size configurable
|
2 mesi fa |
|
Gilad S.
|
3095041e0f
feat: use `build: "autoAttempt"` on `getLlama`
|
2 mesi fa |
|
Big (daocoding)
|
b71649b12d
feat: add QMD_EMBED_MODEL env var for multilingual embedding support
|
2 mesi fa |
|
Tobi Lütke
|
5233e676d9
fix(rerank): truncate documents exceeding 2048-token context size
|
3 mesi fa |
|
Tobias Lütke
|
67e2aab18c
Merge pull request #206 from tobi/liquidai-query-expansion
|
3 mesi fa |
|
Tobi Lütke
|
57f7caa93b
feat: add LiquidAI LFM2 support for query expansion
|
3 mesi fa |
|
Tobi Lutke
|
09803a75b7
feat: compile to JS for npm, release system, full changelog
|
3 mesi fa |
|
Tobi Lütke
|
392934e78a
perf: CPU parallelism via multi-context thread splitting
|
3 mesi fa |
|
Tobi Lütke
|
0a941c442f
perf: flash attention, right-sized contexts, cleaner GPU detection
|
3 mesi fa |
|
Tobi Lütke
|
4ac95b5e26
perf: adaptive parallel contexts for embed + rerank, fix VRAM waste
|
3 mesi fa |
|
Tobi Lütke
|
0a0e1e6f29
perf: parallel reranking with multiple contexts (2.7x speedup)
|
3 mesi fa |
|
Tobi Lütke
|
ee86bba45e
feat: auto-detect GPU acceleration + device info in status
|
3 mesi fa |
|
Tobi Lütke
|
102ff861d3
fix: use Qwen3 recommended sampling params to prevent repetition loops
|
3 mesi fa |
|
Tobi Lütke
|
479b68bbf1
add qmd model pull and refresh logic
|
3 mesi fa |
|
Tobi Lutke
|
7de18ee066
Merge main into finetune
|
3 mesi fa |
|
Tobi Lutke
|
785620467a
refactor: reorder output format to put hyde line first
|
3 mesi fa |
|
Tobi Lütke
|
32d313ad6b
Add LLM session management for lifecycle safety
|
3 mesi fa |
|
Christopher Jones
|
6d9871d2f5
Fix DisposedError during slow batch embedding (#41)
|
3 mesi fa |
|
Tobias Lütke
|
eb1b77c8cb
Deploy fine-tuned GRPO model as default query expansion (#67)
|
3 mesi fa |
|
Tobi Lutke
|
8572c2fd94
Deploy fine-tuned GRPO model as default for query expansion
|
3 mesi fa |
|
Freeman Jiang
|
bfb0eebc3e
fix: use sequential embedding on CPU-only systems to avoid race condition (#54)
|
3 mesi fa |
|
Sergey Gavrilyuk
|
bebee61bec
Fix case sensitivity in Qwen3-1.7B model filename
|
4 mesi fa |
