Public/koboldcpp
1
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-09-20 01:31:42 +02:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'upstream' into concedo_experimental

Browse Source 
# Conflicts:
#	CODEOWNERS
#	docs/ops.md
#	docs/ops/SYCL.csv
#	examples/embedding/README.md
#	ggml/src/ggml-cann/aclnn_ops.cpp
#	ggml/src/ggml-cann/ggml-cann.cpp
#	ggml/src/ggml-sycl/backend.hpp
#	ggml/src/ggml-sycl/ggml-sycl.cpp
#	ggml/src/ggml-sycl/norm.cpp
#	ggml/src/ggml-sycl/norm.hpp
#	scripts/snapdragon/adb/run-bench.sh
#	scripts/snapdragon/adb/run-cli.sh
#	src/llama-batch.cpp
#	tests/test-backend-ops.cpp
#	tests/test-chat.cpp
#	tests/test-json-schema-to-grammar.cpp
#	tools/llama-bench/README.md
This commit is contained in:
Concedo
2025-10-30 13:44:46 +08:00
parent 472438aad3 8b11deea46
commit 16cbe9f24e
55 changed files with 2408 additions and 1222 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tools/server/public_legacy/json-schema-to-grammar.mjs
+8 -3
View File
@@ -345,10 +345,14 @@ export class SchemaConverter {
const selectors = ref.split('#')[1].split('/').slice(1);
for (const sel of selectors) {
if (!target || !(sel in target)) {
const selIndex = parseInt(sel, 10);
if (target && sel in target) {
target = target[sel];
} else if (target && selIndex in target) {
target = target[selIndex];
} else {
throw new Error(`Error resolving ref ${ref}: ${sel} not in ${JSON.stringify(target)}`);
}
target = target[sel];
}
this._refs[ref] = target;
@@ -594,7 +598,8 @@ export class SchemaConverter {
}
_resolveRef(ref) {
let refName = ref.split('/').pop();
let refFragment = ref.split('#').pop();
let refName = 'ref' + refFragment.replace(/[^a-zA-Z0-9-]+/g, '-');
if (!(refName in this._rules) && !this._refsBeingResolved.has(ref)) {
this._refsBeingResolved.add(ref);
const resolved = this._refs[ref];
tools/server/server.cpp
+3 -24
View File
@@ -2866,10 +2866,12 @@ struct server_context {
// if context shifting is disabled, make sure that we don't run out of context
if (!params_base.ctx_shift && slot.n_past + 1 >= slot.n_ctx) {
slot.truncated = true;
slot.stop = STOP_TYPE_LIMIT;
slot.has_next_token = false;
SLT_DBG(slot, "stopped due to running out of context, n_past = %d, n_ctx = %d\n", slot.n_past, slot.n_ctx);
SLT_DBG(slot, "stopped due to running out of context capacity, n_past = %d, n_prompt_tokens = %d, n_decoded = %d, n_ctx = %d\n",
slot.n_decoded, slot.n_prompt_tokens(), slot.n_past, slot.n_ctx);
}
// check the limits
@@ -2929,16 +2931,6 @@ struct server_context {
}
}
// if context shift is disabled, we stop when it reaches the context limit
if (slot.n_past >= slot.n_ctx) {
slot.truncated = true;
slot.stop = STOP_TYPE_LIMIT;
slot.has_next_token = false;
SLT_DBG(slot, "stopped due to running out of context capacity, n_past = %d, n_prompt_tokens = %d, n_decoded = %d, n_ctx = %d\n",
slot.n_decoded, slot.n_prompt_tokens(), slot.n_past, slot.n_ctx);
}
if (llama_vocab_is_eog(vocab, result.tok)) {
slot.stop = STOP_TYPE_EOS;
slot.has_next_token = false;
@@ -2946,19 +2938,6 @@ struct server_context {
SLT_DBG(slot, "%s", "stopped by EOS\n");
}
const auto n_ctx_train = llama_model_n_ctx_train(model);
if (slot.task->params.n_predict < 1 && slot.n_prompt_tokens() + slot.n_decoded >= n_ctx_train) {
slot.truncated = true;
slot.stop = STOP_TYPE_LIMIT;
slot.has_next_token = false; // stop prediction
SLT_WRN(slot,
"n_predict (%d) is set for infinite generation. "
"Limiting generated tokens to n_ctx_train (%d) to avoid EOS-less generation infinite loop\n",
slot.task->params.n_predict, n_ctx_train);
}
SLT_DBG(slot, "n_decoded = %d, n_remaining = %d, next token: %5d '%s'\n", slot.n_decoded, slot.n_remaining, result.tok, token_str.c_str());
return slot.has_next_token; // continue
tools/server/tests/unit/test_ctx_shift.py
+1 -1
View File
@@ -45,7 +45,7 @@ def test_ctx_shift_enabled():
@pytest.mark.parametrize("n_predict,n_token_output,truncated", [
(64, 64, False),
(-1, 120, True),
(-1, 248, True), # 8 tokens prompt + 248 tokens generated = 256 tokens total
])
def test_ctx_shift_disabled_short_prompt(n_predict: int, n_token_output: int, truncated: bool):
global server