Default Branch

021dcf089d · Merge pull request #9824 from ollama/mxyng/sched · Updated 2025-03-17 23:41:37 +01:00

Branches

c75b428249 · fix: fixes a memory leak in bfloat16 package · Updated 2025-03-18 05:46:12 +01:00

0
1

253b3c7a25 · remove errors from sample, add tests · Updated 2025-03-18 04:16:42 +01:00

5
2

f94155fba2 · do not add both consolidated and parts to model · Updated 2025-03-18 00:33:43 +01:00

6
4

22f2f6e229 · llm: set done reason at server level · Updated 2025-03-17 20:01:52 +01:00

6
1

8025781dce · wip · Updated 2025-03-17 18:57:10 +01:00

6
3

c7c751647d · model: support for mistral-small in the ollama runner · Updated 2025-03-15 00:56:39 +01:00

17
1

bc46d0f2dd · use benchmark loop · Updated 2025-03-14 21:52:33 +01:00

26
3

73a1e99f8a · logging: add a new customer logger and trace method · Updated 2025-03-14 00:10:59 +01:00

26
1

a5d638dfe7 · extras · Updated 2025-03-12 21:12:29 +01:00

37
1

f257f1fd04 · sample: do all sorting in topK · Updated 2025-03-12 19:20:18 +01:00

44
3

9622b928b4 · extras · Updated 2025-03-12 18:28:59 +01:00

44
4

15b8875cfc · runner/ollamarunner: set temperature to 0 when images are provided · Updated 2025-03-11 23:41:18 +01:00

44
1

9c23f11850 · pr feedback · Updated 2025-03-11 05:24:06 +01:00

84
8

12a8b00b34 · server: allow running embed models in parallel · Updated 2025-03-10 21:34:09 +01:00

84
1

81888abbe4 · wip: apply gbnf vocab to logits · Updated 2025-03-07 06:44:52 +01:00

115
1

554aed43bd · llama: add patch to avoid errors on arrays of strigns · Updated 2025-03-06 02:42:11 +01:00

113
1

b48b6f85cd · server/internal/client/ollama: hold DiskCache on Registry · Updated 2025-03-03 00:43:24 +01:00

130
1

3ea20883a8 · ci: use separate, faster runners for windows CI · Updated 2025-03-01 23:37:28 +01:00

137
1

9c1204b686 · server/internal/internal/names: validate names · Updated 2025-03-01 01:30:42 +01:00

140
1

5beede47d9 · ml: Add support for quantized KV cache · Updated 2025-02-28 02:09:16 +01:00

148
4