mirror of
https://github.com/kvcache-ai/ktransformers.git
synced 2026-05-05 23:50:14 +00:00
* refactor: move legacy code to archive/ directory - Moved ktransformers, csrc, third_party, merge_tensors to archive/ - Moved build scripts and configurations to archive/ - Kept kt-kernel, KT-SFT, doc, and README files in root - Preserved complete git history for all moved files * refactor: restructure repository to focus on kt-kernel and KT-SFT modules * fix README * fix README * fix README * fix README * docs: add performance benchmarks to kt-kernel section Add comprehensive performance data for kt-kernel to match KT-SFT's presentation: - AMX kernel optimization: 21.3 TFLOPS (3.9× faster than PyTorch) - Prefill phase: up to 20× speedup vs baseline - Decode phase: up to 4× speedup - NUMA optimization: up to 63% throughput improvement - Multi-GPU (8×L20): 227.85 tokens/s total throughput with DeepSeek-R1 FP8 Source: https://lmsys.org/blog/2025-10-22-KTransformers/ This provides users with concrete performance metrics for both core modules, making it easier to understand the capabilities of each component. * refactor: improve kt-kernel performance data with specific hardware and models Replace generic performance descriptions with concrete benchmarks: - Specify exact hardware: 8×L20 GPU + Xeon Gold 6454S, Single/Dual-socket Xeon + AMX - Include specific models: DeepSeek-R1-0528 (FP8), DeepSeek-V3 (671B) - Show detailed metrics: total throughput, output throughput, concurrency details - Match KT-SFT presentation style for consistency This provides users with actionable performance data they can use to evaluate hardware requirements and expected performance for their use cases. * fix README * docs: clean up performance table and improve formatting * add pic for README * refactor: simplify .gitmodules and backup legacy submodules - Remove 7 legacy submodules from root .gitmodules (archive/third_party/*) - Keep only 2 active submodules for kt-kernel (llama.cpp, pybind11) - Backup complete .gitmodules to archive/.gitmodules - Add documentation in archive/README.md for researchers who need legacy submodules This reduces initial clone size by ~500MB and avoids downloading unused dependencies. * refactor: move doc/ back to root directory Keep documentation in root for easier access and maintenance. * refactor: consolidate all images to doc/assets/ - Move kt-kernel/assets/heterogeneous_computing.png to doc/assets/ - Remove KT-SFT/assets/ (images already in doc/assets/) - Update KT-SFT/README.md image references to ../doc/assets/ - Eliminates ~7.9MB image duplication - Centralizes all documentation assets in one location * fix pic path for README
79 lines
2.4 KiB
Python
79 lines
2.4 KiB
Python
import time
|
|
|
|
|
|
def format_time(seconds):
|
|
units = [
|
|
("hours", 3600),
|
|
("minutes", 60),
|
|
("seconds", 1),
|
|
("milliseconds", 1e-3),
|
|
("microseconds", 1e-6),
|
|
]
|
|
|
|
for unit_name, unit_value in units:
|
|
if seconds >= unit_value:
|
|
time_value = seconds / unit_value
|
|
return f"{time_value:.2f} {unit_name}"
|
|
return "0 seconds" # Handle case for 0 seconds
|
|
|
|
|
|
class Profiler:
|
|
def __init__(self):
|
|
self.timers = {}
|
|
self.counters = {}
|
|
|
|
def create_timer(self, name):
|
|
self.timers[name] = {
|
|
"start_time": None,
|
|
"elapsed_time": 0,
|
|
"running": False,
|
|
}
|
|
|
|
def start_timer(self, name):
|
|
if name not in self.timers:
|
|
raise ValueError(f"Timer '{name}' does not exist.")
|
|
if self.timers[name]["running"]:
|
|
raise ValueError(f"Timer '{name}' is already running.")
|
|
self.timers[name]["start_time"] = time.time()
|
|
self.timers[name]["running"] = True
|
|
|
|
def pause_timer(self, name):
|
|
if name not in self.timers:
|
|
raise ValueError(f"Timer '{name}' does not exist.")
|
|
if not self.timers[name]["running"]:
|
|
raise ValueError(f"Timer '{name}' is not running.")
|
|
self.timers[name]["elapsed_time"] += time.time() - self.timers[name]["start_time"]
|
|
self.timers[name]["running"] = False
|
|
|
|
def get_timer_sec(self, name):
|
|
if name not in self.timers:
|
|
raise ValueError(f"Timer '{name}' does not exist.")
|
|
if self.timers[name]["running"]:
|
|
current_time = self.timers[name]["elapsed_time"] + (time.time() - self.timers[name]["start_time"])
|
|
else:
|
|
current_time = self.timers[name]["elapsed_time"]
|
|
return current_time
|
|
|
|
def get_all_timers(self):
|
|
all_timers = {}
|
|
for name in self.timers:
|
|
all_timers[name] = self.get_timer_sec(name)
|
|
return all_timers
|
|
|
|
def report_timer_string(self, name):
|
|
return f"{name} elapsed time: {format_time(self.get_timer_sec(name))}"
|
|
|
|
def create_and_start_timer(self, name):
|
|
self.create_timer(name)
|
|
self.start_timer(name)
|
|
|
|
|
|
# Counter
|
|
def inc(self,key:str,delta:int=1):
|
|
self.counters[key] = self.counters.get(key,0) + delta
|
|
|
|
def set_counter(self,key:str,to=0):
|
|
self.counters[key] = to
|
|
|
|
def get_counter(self,key:str):
|
|
return self.counters.get(key,0)
|