mirror of
https://github.com/dogkeeper886/ollama37.git
synced 2025-12-11 00:07:07 +00:00
ef14fb5b26
This commit represents a complete rework after pulling the latest changes from official ollama/ollama repository and re-applying Tesla K80 compatibility patches. ## Key Changes ### CUDA Compute Capability 3.7 Support (Tesla K80) - Added sm_37 (compute 3.7) to CMAKE_CUDA_ARCHITECTURES in CMakeLists.txt - Updated CMakePresets.json to include compute 3.7 in "CUDA 11" preset - Using 37-virtual (PTX with JIT compilation) for maximum compatibility ### Legacy Toolchain Compatibility - **NVIDIA Driver**: 470.256.02 (last version supporting Kepler/K80) - **CUDA Version**: 11.4.4 (last CUDA 11.x supporting compute 3.7) - **GCC Version**: 10.5.0 (required by CUDA 11.4 host_config.h) ### CPU Architecture Trade-offs Due to GCC 10.5 limitation, sacrificed newer CPU optimizations: - Alderlake CPU variant enabled WITHOUT AVX_VNNI (requires GCC 11+) - Still supports: SSE4.2, AVX, F16C, AVX2, BMI2, FMA - Performance impact: ~3-7% on newer CPUs (acceptable for K80 compatibility) ### Build System Updates - Modified ml/backend/ggml/ggml/src/ggml-cuda/CMakeLists.txt for compute 3.7 - Added -Wno-deprecated-gpu-targets flag to suppress warnings - Updated ml/backend/ggml/ggml/src/CMakeLists.txt for Alderlake without AVX_VNNI ### Upstream Sync Merged latest llama.cpp changes including: - Enhanced KV cache management with ISWA and hybrid memory support - Improved multi-modal support (mtmd framework) - New model architectures (Gemma3, Llama4, Qwen3, etc.) - GPU backend improvements for CUDA, Metal, and ROCm - Updated quantization support and GGUF format handling ### Documentation - Updated CLAUDE.md with comprehensive build instructions - Documented toolchain constraints and CPU architecture trade-offs - Removed outdated CI/CD workflows (tesla-k80-*.yml) - Cleaned up temporary development artifacts ## Rationale This fork maintains Tesla K80 GPU support (compute 3.7) which was dropped in official Ollama due to legacy driver/CUDA requirements. The toolchain constraint creates a deadlock: - K80 → Driver 470 → CUDA 11.4 → GCC 10 → No AVX_VNNI We accept the loss of cutting-edge CPU optimizations to enable running modern LLMs on legacy but still capable Tesla K80 hardware (12GB VRAM per GPU). 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
83 lines
2.6 KiB
Go
83 lines
2.6 KiB
Go
package llamarunner
|
|
|
|
import (
|
|
"reflect"
|
|
"testing"
|
|
|
|
"github.com/ollama/ollama/llama"
|
|
)
|
|
|
|
func TestImageCache(t *testing.T) {
|
|
cache := ImageContext{images: make([]imageCache, 4)}
|
|
|
|
valA := []llama.MtmdChunk{{Embed: []float32{0.1, 0.2}}, {Embed: []float32{0.3}}}
|
|
valB := []llama.MtmdChunk{{Embed: []float32{0.4}}, {Embed: []float32{0.5}}, {Embed: []float32{0.6}}}
|
|
valC := []llama.MtmdChunk{{Embed: []float32{0.7}}}
|
|
valD := []llama.MtmdChunk{{Embed: []float32{0.8}}}
|
|
valE := []llama.MtmdChunk{{Embed: []float32{0.9}}}
|
|
|
|
// Empty cache
|
|
result, err := cache.findImage(0x5adb61d31933a946)
|
|
if err != errImageNotFound {
|
|
t.Errorf("found result in empty cache: result %v, err %v", result, err)
|
|
}
|
|
|
|
// Insert A
|
|
cache.addImage(0x5adb61d31933a946, valA)
|
|
|
|
result, err = cache.findImage(0x5adb61d31933a946)
|
|
if !reflect.DeepEqual(result, valA) {
|
|
t.Errorf("failed to find expected value: result %v, err %v", result, err)
|
|
}
|
|
|
|
// Insert B
|
|
cache.addImage(0x011551369a34a901, valB)
|
|
|
|
result, err = cache.findImage(0x5adb61d31933a946)
|
|
if !reflect.DeepEqual(result, valA) {
|
|
t.Errorf("failed to find expected value: result %v, err %v", result, err)
|
|
}
|
|
result, err = cache.findImage(0x011551369a34a901)
|
|
if !reflect.DeepEqual(result, valB) {
|
|
t.Errorf("failed to find expected value: result %v, err %v", result, err)
|
|
}
|
|
|
|
// Replace B with C
|
|
cache.addImage(0x011551369a34a901, valC)
|
|
|
|
result, err = cache.findImage(0x5adb61d31933a946)
|
|
if !reflect.DeepEqual(result, valA) {
|
|
t.Errorf("failed to find expected value: result %v, err %v", result, err)
|
|
}
|
|
result, err = cache.findImage(0x011551369a34a901)
|
|
if !reflect.DeepEqual(result, valC) {
|
|
t.Errorf("failed to find expected value: result %v, err %v", result, err)
|
|
}
|
|
|
|
// Evict A
|
|
cache.addImage(0x756b218a517e7353, valB)
|
|
cache.addImage(0x75e5e8d35d7e3967, valD)
|
|
cache.addImage(0xd96f7f268ca0646e, valE)
|
|
|
|
result, err = cache.findImage(0x5adb61d31933a946)
|
|
if reflect.DeepEqual(result, valA) {
|
|
t.Errorf("failed to find expected value: result %v, err %v", result, err)
|
|
}
|
|
result, err = cache.findImage(0x756b218a517e7353)
|
|
if !reflect.DeepEqual(result, valB) {
|
|
t.Errorf("failed to find expected value: result %v, err %v", result, err)
|
|
}
|
|
result, err = cache.findImage(0x011551369a34a901)
|
|
if !reflect.DeepEqual(result, valC) {
|
|
t.Errorf("failed to find expected value: result %v, err %v", result, err)
|
|
}
|
|
result, err = cache.findImage(0x75e5e8d35d7e3967)
|
|
if !reflect.DeepEqual(result, valD) {
|
|
t.Errorf("failed to find expected value: result %v, err %v", result, err)
|
|
}
|
|
result, err = cache.findImage(0xd96f7f268ca0646e)
|
|
if !reflect.DeepEqual(result, valE) {
|
|
t.Errorf("failed to find expected value: result %v, err %v", result, err)
|
|
}
|
|
}
|