refactor: benchmarks (#33896)

* refactor: benchmarks

Based on a discussion with @LysandreJik & @ArthurZucker, the goal of
this PR is to improve transformers' benchmark system.

This is a WIP, for the moment the infrastructure required to make things
work is not ready. Will update the PR description when it is the case.

* feat: add db init in benchmarks CI

* fix: pg_config is missing in runner

* fix: add psql to the runner

* fix: connect info from env vars + PR comments

* refactor: set database as env var

* fix: invalid working directory

* fix: `commit_msg` -> `commit_message`

* fix: git marking checked out repo as unsafe

* feat: add logging

* fix: invalid device

* feat: update grafana dashboard for prod grafana

* feat: add `commit_id` to header table

* feat: commit latest version of dashboard

* feat: move measurements into json field

* feat: remove drop table migration queries

* fix: `torch.arrange` -> `torch.arange`

* fix: add missing `s` to `cache_position` positional argument

* fix: change model

* revert: `cache_positions` -> `cache_position`

* fix: set device for `StaticCache`

* fix: set `StaticCache` dtype

* feat: limit max cache len

* fix script

* raise error on failure!

* not try catch

* try to skip generate compilation

* update

* update docker image!

* update

* update again!@

* update

* updates

* ???

* ??

* use `torch.cuda.synchronize()`

* fix json

* nits

* fix

* fixed!

* f**k

* feat: add TTNT panels

* feat: add try except

---------

Co-authored-by: Arthur Zucker <arthur.zucker@gmail.com>

.github/workflows/benchmark.yml

@@ -1,43 +1,72 @@
 name: Self-hosted runner (benchmark)
 
 on:
-  schedule:
-    - cron: "17 2 * * *"
-  workflow_call:
+  push:
+    branches: [main]
+  pull_request:
+    types: [ opened, labeled, reopened, synchronize ]
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
+  cancel-in-progress: true
 
 env:
   HF_HOME: /mnt/cache
-  TF_FORCE_GPU_ALLOW_GROWTH: true
-
 
 jobs:
   benchmark:
     name: Benchmark
-    runs-on: 
+    runs-on:
       group: aws-g5-4xlarge-cache
     container:
-      image: huggingface/transformers-all-latest-gpu
-      options: --gpus all --privileged --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
+      image: huggingface/transformers-pytorch-gpu
+      options: --gpus all --privileged --ipc host
     steps:
-      - name: Update clone
-        working-directory: /transformers
+      - name: Get repo
+        if: github.event_name == 'pull_request'
+        uses: actions/checkout@v4
+        with:
+          ref: ${{ github.event.pull_request.head.sha }}
+
+      - name: Get repo
+        if: github.event_name == 'push'
+        uses: actions/checkout@v4
+        with:
+          ref: ${{ github.sha }}
+
+      - name: Install libpq-dev & psql
         run: |
-          git fetch && git checkout ${{ github.sha }}
+          apt update
+          apt install -y libpq-dev postgresql-client
+
+      - name: Install benchmark script dependencies
+        run: python3 -m pip install -r benchmark/requirements.txt
 
       - name: Reinstall transformers in edit mode (remove the one installed during docker image build)
         working-directory: /transformers
-        run: python3 -m pip uninstall -y transformers && python3 -m pip install -e .
+        run: python3 -m pip uninstall -y transformers && python3 -m pip install -e ".[torch]"
 
-      - name: Benchmark (daily)
-        if: github.event_name == 'schedule'
-        working-directory: /transformers
+      - name: Run database init script
         run: |
-          python3 -m pip install optimum-benchmark>=0.3.0
-          HF_TOKEN=${{ secrets.TRANSFORMERS_BENCHMARK_TOKEN }} python3 benchmark/benchmark.py --repo_id hf-internal-testing/benchmark_results --path_in_repo $(date +'%Y-%m-%d') --config-dir benchmark/config --config-name generation --commit=${{ github.sha }} backend.model=google/gemma-2b backend.cache_implementation=null,static backend.torch_compile=false,true --multirun
+          psql -f benchmark/init_db.sql
+        env:
+          PGDATABASE: metrics
+          PGHOST: ${{ secrets.TRANSFORMERS_BENCHMARKS_PGHOST }}
+          PGUSER: transformers_benchmarks
+          PGPASSWORD: ${{ secrets.TRANSFORMERS_BENCHMARKS_PGPASSWORD }}
 
-      - name: Benchmark (merged to main event)
-        if: github.event_name == 'push' && github.ref_name == 'main'
-        working-directory: /transformers
+      - name: Run benchmark
         run: |
-          python3 -m pip install optimum-benchmark>=0.3.0
-          HF_TOKEN=${{ secrets.TRANSFORMERS_BENCHMARK_TOKEN }} python3 benchmark/benchmark.py --repo_id hf-internal-testing/benchmark_results_merge_event --path_in_repo $(date +'%Y-%m-%d') --config-dir benchmark/config --config-name generation --commit=${{ github.sha }} backend.model=google/gemma-2b backend.cache_implementation=null,static backend.torch_compile=false,true --multirun
+          git config --global --add safe.directory /__w/transformers/transformers
+          if [ "$GITHUB_EVENT_NAME" = "pull_request" ]; then
+            commit_id=$(echo "${{ github.event.pull_request.head.sha }}")
+          elif [ "$GITHUB_EVENT_NAME" = "push" ]; then
+            commit_id=$GITHUB_SHA
+          fi
+          commit_msg=$(git show -s --format=%s | cut -c1-70)
+          python3 benchmark/llama.py "${{ github.head_ref || github.ref_name }}" "$commit_id" "$commit_msg"
+        env:
+          HF_TOKEN: ${{ secrets.HF_HUB_READ_TOKEN }}
+          PGHOST: ${{ secrets.TRANSFORMERS_BENCHMARKS_PGHOST }}
+          PGUSER: transformers_benchmarks
+          PGPASSWORD: ${{ secrets.TRANSFORMERS_BENCHMARKS_PGPASSWORD }}

benchmark/init_db.sql

@@ -0,0 +1,26 @@
+CREATE TABLE IF NOT EXISTS benchmarks (
+  benchmark_id SERIAL PRIMARY KEY,
+  branch VARCHAR(255),
+  commit_id VARCHAR(72),
+  commit_message VARCHAR(70),
+  gpu_name VARCHAR(255),
+  created_at timestamp without time zone NOT NULL DEFAULT (current_timestamp AT TIME ZONE 'UTC')
+);
+
+CREATE TABLE IF NOT EXISTS device_measurements (
+  measurement_id SERIAL PRIMARY KEY,
+  benchmark_id int REFERENCES benchmarks (benchmark_id),
+  cpu_util double precision,
+  mem_megabytes double precision,
+  gpu_util double precision,
+  gpu_mem_megabytes double precision,
+  time timestamp without time zone NOT NULL DEFAULT (current_timestamp AT TIME ZONE 'UTC')
+);
+
+CREATE TABLE IF NOT EXISTS model_measurements (
+  measurement_id SERIAL PRIMARY KEY,
+  benchmark_id int REFERENCES benchmarks (benchmark_id),
+  measurements jsonb,
+  time timestamp without time zone NOT NULL DEFAULT (current_timestamp AT TIME ZONE 'UTC')
+);
+

benchmark/llama.py

@@ -0,0 +1,404 @@
+import argparse
+import json
+import logging
+import os
+import sys
+from statistics import mean
+from threading import Event, Thread
+from time import perf_counter, sleep
+from typing import Optional
+import gpustat
+import psutil
+import psycopg2
+import torch
+
+from transformers import AutoModelForCausalLM, AutoTokenizer, GenerationConfig, StaticCache
+from psycopg2.extras import Json
+from psycopg2.extensions import register_adapter
+
+
+os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1"
+
+logger = logging.getLogger(__name__)
+logger.setLevel(logging.INFO)
+
+handler = logging.StreamHandler(sys.stdout)
+handler.setLevel(logging.INFO)
+formatter = logging.Formatter("[%(levelname)s - %(asctime)s] %(message)s")
+handler.setFormatter(formatter)
+logger.addHandler(handler)
+
+os.environ["TOKENIZERS_PARALLELISM"] = "1"
+torch.set_float32_matmul_precision("high")
+register_adapter(dict, Json)
+
+
+def parse_arguments():
+    """
+    Parse command line arguments for the benchmarking CLI.
+    """
+    parser = argparse.ArgumentParser(description="CLI for benchmarking the huggingface/transformers.")
+
+    parser.add_argument(
+        "branch",
+        type=str,
+        help="The branch name on which the benchmarking is performed.",
+    )
+
+    parser.add_argument(
+        "commit_id",
+        type=str,
+        help="The commit hash on which the benchmarking is performed.",
+    )
+
+    parser.add_argument(
+        "commit_msg",
+        type=str,
+        help="The commit message associated with the commit, truncated to 70 characters.",
+    )
+
+    args = parser.parse_args()
+
+    return args.branch, args.commit_id, args.commit_msg
+
+
+def collect_metrics(benchmark_id, continue_metric_collection):
+    p = psutil.Process(os.getpid())
+    conn = psycopg2.connect("dbname=metrics")
+    cur = conn.cursor()
+    while not continue_metric_collection.is_set():
+        with p.oneshot():
+            cpu_util = p.cpu_percent()
+            mem_megabytes = p.memory_info().rss / (1024 * 1024)
+        gpu_stats = gpustat.GPUStatCollection.new_query()
+        gpu_util = gpu_stats[0]["utilization.gpu"]
+        gpu_mem_megabytes = gpu_stats[0]["memory.used"]
+        cur.execute(
+            "INSERT INTO device_measurements (benchmark_id, cpu_util, mem_megabytes, gpu_util, gpu_mem_megabytes) VALUES (%s, %s, %s, %s, %s)",
+            (benchmark_id, cpu_util, mem_megabytes, gpu_util, gpu_mem_megabytes),
+        )
+        sleep(0.01)
+        conn.commit()
+    conn.close()
+
+
+def run_benchmark(branch: str, commit_id: str, commit_msg: str, num_tokens_to_generate=100):
+    continue_metric_collection = Event()
+    metrics_thread = None
+    try:
+        gpu_stats = gpustat.GPUStatCollection.new_query()
+        gpu_name = gpu_stats[0]["name"]
+        conn = psycopg2.connect("dbname=metrics")
+        cur = conn.cursor()
+        cur.execute(
+            "INSERT INTO benchmarks (branch, commit_id, commit_message, gpu_name) VALUES (%s, %s, %s, %s) RETURNING benchmark_id",
+            (branch, commit_id, commit_msg, gpu_name),
+        )
+        conn.commit()
+        benchmark_id = cur.fetchone()[0]
+        metrics_thread = Thread(target=collect_metrics, args=[benchmark_id, continue_metric_collection])
+        metrics_thread.start()
+
+        os.environ["TOKENIZERS_PARALLELISM"] = "false"  # silence warnings when compiling
+
+        device = "cuda"
+        ckpt = "meta-llama/Llama-2-7b-hf"
+
+        # This is to avoid counting download in model load time measurement
+        model = AutoModelForCausalLM.from_pretrained(ckpt, torch_dtype=torch.float16)
+        gen_config = GenerationConfig(do_sample=False, top_p=1, temperature=1)
+        start = perf_counter()
+        model = AutoModelForCausalLM.from_pretrained(
+            ckpt, torch_dtype=torch.float16, generation_config=gen_config
+        ).eval()
+        model.to(device)
+        torch.cuda.synchronize()
+        end = perf_counter()
+        model_load_time = end - start
+        logger.info(f"loaded model in: {model_load_time}s")
+
+        tokenizer = AutoTokenizer.from_pretrained(ckpt)
+
+        prompt = "Why dogs are so cute?"
+        inputs = tokenizer(prompt, return_tensors="pt").to(device)
+
+        # Specify the max length (including both the prompt and the response)
+        # When calling `generate` with `cache_implementation="static" later, this is also used to create a `StaticCache` object
+        # with sequence length = `max_length`. The longer the more you will re-use it
+        seq_length = inputs["input_ids"].shape[1]
+        model.generation_config.max_length = seq_length + num_tokens_to_generate
+        batch_size = inputs["input_ids"].shape[0]
+
+        # Copied from the gpt-fast repo
+        def multinomial_sample_one_no_sync(probs_sort):  # Does multinomial sampling without a cuda synchronization
+            q = torch.empty_like(probs_sort).exponential_(1)
+            return torch.argmax(probs_sort / q, dim=-1, keepdim=True).to(dtype=torch.int)
+
+        def logits_to_probs(logits, temperature: float = 1.0, top_k: Optional[int] = None):
+            logits = logits / max(temperature, 1e-5)
+
+            if top_k is not None:
+                v, _ = torch.topk(logits, min(top_k, logits.size(-1)))
+                pivot = v.select(-1, -1).unsqueeze(-1)
+                logits = torch.where(logits < pivot, -float("Inf"), logits)
+            probs = torch.nn.functional.softmax(logits, dim=-1)
+            return probs
+
+        def sample(logits, temperature: float = 1.0, top_k: Optional[int] = None):
+            probs = logits_to_probs(logits[:, -1], temperature, top_k)
+            idx_next = multinomial_sample_one_no_sync(probs)
+            return idx_next, probs
+
+        def decode_one_token(model, cur_token, cache_position, past_key_values):
+            logits = model(
+                cur_token,
+                cache_position=cache_position,
+                past_key_values=past_key_values,
+                return_dict=False,
+                use_cache=True,
+            )[0]
+            new_token = sample(logits, temperature=0.6, top_k=5)[0]
+            return new_token
+
+        #########
+        # Eager #
+        #########
+        with torch.no_grad():
+            past_key_values = StaticCache(
+                model.config,
+                batch_size=batch_size,
+                device=device,
+                dtype=torch.float16,
+                max_cache_len=seq_length + num_tokens_to_generate,
+            )
+            cache_position = torch.arange(seq_length, device=device)
+            start = perf_counter()
+            model(
+                **inputs,
+                cache_position=cache_position,
+                past_key_values=past_key_values,
+                return_dict=False,
+                use_cache=True,
+            )
+            end = perf_counter()
+            first_eager_fwd_pass_time = end - start
+            logger.info(f"completed first eager fwd pass in: {first_eager_fwd_pass_time}s")
+            start = perf_counter()
+            output = model.generate(**inputs, do_sample=False)
+            end = perf_counter()
+            first_eager_generate_time = end - start
+            logger.info(f"completed first eager generation in: {first_eager_generate_time}s")
+            logger.info(f"generated: {tokenizer.batch_decode(output.cpu().tolist())}")
+
+            past_key_values = StaticCache(
+                model.config,
+                batch_size=batch_size,
+                device=device,
+                dtype=torch.float16,
+                max_cache_len=seq_length + num_tokens_to_generate,
+            )
+            cache_position = torch.arange(seq_length, device=device)
+            start = perf_counter()
+            model(
+                **inputs,
+                cache_position=cache_position,
+                past_key_values=past_key_values,
+                return_dict=False,
+                use_cache=True,
+            )
+            end = perf_counter()
+            second_eager_fwd_pass_time = end - start
+            logger.info(f"completed second eager fwd pass in: {second_eager_fwd_pass_time}s")
+            start = perf_counter()
+            model.generate(**inputs, do_sample=False)
+            end = perf_counter()
+            second_eager_generate_time = end - start
+            logger.info(f"completed second eager generation in: {second_eager_generate_time}s")
+            logger.info(f"generated: {tokenizer.batch_decode(output.cpu().tolist())}")
+
+            torch.compiler.reset()
+
+            ################
+            # Forward pass #
+            ################
+
+            # `torch.compile(model, ...)` is not recommended as you compile callbacks
+            # and full generate. We recommend compiling only the forward for now.
+            # "reduce-overhead" will use cudagraphs.
+            generated_ids = torch.zeros(
+                (batch_size, num_tokens_to_generate + seq_length), dtype=torch.int, device=device
+            )
+
+            generated_ids[:, :seq_length] = inputs["input_ids"]
+            decode_one_token = torch.compile(decode_one_token, mode="reduce-overhead", fullgraph=True)
+            # model.forward = torch.compile(model.forward, mode="reduce-overhead", fullgraph=True)
+            # TODO use  decode_one_token(model, input_id.clone(), cache_position) for verification
+            past_key_values = StaticCache(
+                model.config,
+                batch_size=batch_size,
+                device=device,
+                dtype=torch.float16,
+                max_cache_len=seq_length + num_tokens_to_generate + 10,
+            )
+            cache_position = torch.arange(seq_length, device=device)
+            all_generated_tokens = []
+            ### First compile, prefill
+            start = perf_counter()
+            next_token = decode_one_token(
+                model, inputs["input_ids"], cache_position=cache_position, past_key_values=past_key_values
+            )
+            torch.cuda.synchronize()
+            end = perf_counter()
+            time_to_first_token = end - start
+            logger.info(f"completed first compile generation in: {time_to_first_token}s")
+            cache_position += 1
+            all_generated_tokens += next_token.clone().detach().cpu().tolist()
+
+            cache_position = torch.tensor([seq_length], device=device)
+            ### First compile, decoding
+            start = perf_counter()
+            next_token = decode_one_token(
+                model, next_token.clone(), cache_position=cache_position, past_key_values=past_key_values
+            )
+            torch.cuda.synchronize()
+            end = perf_counter()
+            time_to_second_token = end - start
+            logger.info(f"completed second compile generation in: {time_to_first_token}s")
+            cache_position += 1
+            all_generated_tokens += next_token.clone().detach().cpu().tolist()
+
+            ### Second compile, decoding
+            start = perf_counter()
+            next_token = decode_one_token(
+                model, next_token.clone(), cache_position=cache_position, past_key_values=past_key_values
+            )
+            torch.cuda.synchronize()
+            end = perf_counter()
+            time_to_third_token = end - start
+            logger.info(f"completed third compile forward in: {time_to_first_token}s")
+            cache_position += 1
+            all_generated_tokens += next_token.clone().detach().cpu().tolist()
+
+            ### Using cuda graphs decoding
+
+            start = perf_counter()
+            for _ in range(1, num_tokens_to_generate):
+                all_generated_tokens += next_token.clone().detach().cpu().tolist()
+                next_token = decode_one_token(
+                    model, next_token.clone(), cache_position=cache_position, past_key_values=past_key_values
+                )
+                cache_position += 1
+            torch.cuda.synchronize()
+            end = perf_counter()
+            mean_time_to_next_token = (end - start) / num_tokens_to_generate
+            logger.info(f"completed next compile generation in: {mean_time_to_next_token}s")
+            logger.info(f"generated: {tokenizer.batch_decode(all_generated_tokens)}")
+
+            ####################
+            # Generate compile #
+            ####################
+            torch.compiler.reset()
+            # we will not compile full generate as it' s to intensive, tho we measure full forward!
+
+            past_key_values = StaticCache(
+                model.config,
+                batch_size=batch_size,
+                device=device,
+                dtype=torch.float16,
+                max_cache_len=seq_length + 128,
+            )
+
+            # 1st call
+            start = perf_counter()
+            output = model.generate(**inputs, past_key_values=past_key_values)
+            torch.cuda.synchronize()
+            end = perf_counter()
+            first_compile_generate_time = end - start
+            logger.info(f"completed first compile generation in: {first_compile_generate_time}s")
+            logger.info(f"generated: {tokenizer.batch_decode(output.cpu().tolist())}")
+
+            past_key_values = StaticCache(
+                model.config,
+                batch_size=batch_size,
+                device=device,
+                dtype=torch.float16,
+                max_cache_len=seq_length + 128,
+            )
+            # 2nd call
+            start = perf_counter()
+            output = model.generate(**inputs, past_key_values=past_key_values)
+            torch.cuda.synchronize()
+            end = perf_counter()
+            second_compile_generate_time = end - start
+            logger.info(f"completed second compile generation in: {second_compile_generate_time}s")
+            logger.info(f"generated: {tokenizer.batch_decode(output.cpu().tolist())}")
+
+            past_key_values = StaticCache(
+                model.config,
+                batch_size=batch_size,
+                device=device,
+                dtype=torch.float16,
+                max_cache_len=seq_length + 128,
+            )
+
+            # 3nd call
+            start = perf_counter()
+            output = model.generate(**inputs, past_key_values=past_key_values)
+            end = perf_counter()
+            third_compile_generate_time = end - start
+            logger.info(f"completed second compile generation in: {third_compile_generate_time}s")
+            logger.info(f"generated: {tokenizer.batch_decode(output.cpu().tolist())}")
+
+            past_key_values = StaticCache(
+                model.config,
+                batch_size=batch_size,
+                device=device,
+                dtype=torch.float16,
+                max_cache_len=seq_length + 128,
+            )
+            # 4th call
+            start = perf_counter()
+            output = model.generate(**inputs, past_key_values=past_key_values)
+            end = perf_counter()
+            fourth_compile_generate_time = end - start
+            logger.info(f"completed second compile generation in: {fourth_compile_generate_time}s")
+            logger.info(f"generated: {tokenizer.batch_decode(output.cpu().tolist())}")
+
+        cur.execute(
+            """
+            INSERT INTO model_measurements (
+                benchmark_id,
+                measurements
+            ) VALUES (%s, %s)
+            """,
+            (
+                benchmark_id,
+                {
+                    "model_load_time": model_load_time,
+                    "first_eager_forward_pass_time_secs": first_eager_fwd_pass_time,
+                    "second_eager_forward_pass_time_secs": second_eager_fwd_pass_time,
+                    "first_eager_generate_time_secs": first_eager_generate_time,
+                    "second_eager_generate_time_secs": second_eager_generate_time,
+                    "time_to_first_token_secs": time_to_first_token,
+                    "time_to_second_token_secs": time_to_second_token,
+                    "time_to_third_token_secs": time_to_third_token,
+                    "time_to_next_token_mean_secs": mean_time_to_next_token,
+                    "first_compile_generate_time_secs": first_compile_generate_time,
+                    "second_compile_generate_time_secs": second_compile_generate_time,
+                    "third_compile_generate_time_secs": third_compile_generate_time,
+                    "fourth_compile_generate_time_secs": fourth_compile_generate_time,
+                },
+            ),
+        )
+        conn.commit()
+        conn.close()
+    except Exception as e:
+        logger.error(f"Caught exception: {e}")
+    continue_metric_collection.set()
+    if metrics_thread is not None:
+        metrics_thread.join()
+
+
+if __name__ == "__main__":
+    branch, commit_id, commit_msg = parse_arguments()
+    run_benchmark(branch, commit_id, commit_msg, num_tokens_to_generate=20)

benchmark/requirements.txt

@@ -0,0 +1,5 @@
+gpustat==1.1.1
+psutil==6.0.0
+psycopg2==2.9.9
+torch>=2.4.0
+hf_transfer