mirrors/ollama-for-amd
1
0
Fork 0
mirror of https://github.com/likelovewant/ollama-for-amd.git synced 2025-12-21 22:33:56 +00:00
Code Issues Packages Projects Releases Wiki Activity

embeddings: modified batch size (#13429)

Browse Source 
This PR detects embedding models and sets batch_size = context_size so the full input fits in a single batch.
Previously, if batch size was smaller than the input, tokens could be split across batches and cause a SIGTRAP crash.
This change ensures all tokens stay in one batch and prevents crashes.
Fixes: #12938 #13054

Co-authored-by: Jesse Gross <jesse@ollama.com>
This commit is contained in:
nicole pardal
2025-12-11 15:36:31 -08:00
committed by GitHub
parent 48e78e9be1
commit 3475d915cb
5 changed files with 78 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
llama/llama.go
View File

@@ -121,7 +121,8 @@ type ContextParams struct {
func NewContextParams(numCtx int, batchSize int, numSeqMax int, threads int, flashAttention bool, kvCacheType string) ContextParams {
params := C.llama_context_default_params()
params.n_ctx = C.uint(numCtx)
params.n_batch = C.uint(batchSize)
params.n_batch = C.uint(batchSize * numSeqMax)
params.n_ubatch = C.uint(batchSize)
params.n_seq_max = C.uint(numSeqMax)
params.n_threads = C.int(threads)
params.n_threads_batch = params.n_threads