[examples] Reports

paice-v1-10-g1243bba/NVIDIA_A100_80GB_PCIe/README.md

@@ -0,0 +1,38 @@
+```
+=================
+Benchmark results
+=================
+                         fail n       perf   sem%   std% peak_memory          score weight
+bert-fp16                   0 1     149.82   0.1%   0.4%       24616     149.818434   0.00
+bert-fp32                   0 1      27.22   0.0%   0.1%       31580      27.217099   0.00
+bert-tf32                   0 1     116.65   0.1%   0.5%       31582     116.647454   0.00
+bert-tf32-fp16              0 1     150.16   0.1%   0.4%       24616     150.156416   3.00
+bf16                        0 1     270.83   0.1%   0.8%        1804     270.832982   0.00
+convnext_large-fp16         0 1     310.32   1.9%  10.0%       27478     310.322641   0.00
+convnext_large-fp32         0 1      42.56   2.2%  11.7%       49598      42.558568   0.00
+convnext_large-tf32         0 1     124.27   4.0%  21.4%       49598     124.274883   0.00
+convnext_large-tf32-fp16    0 1     309.84   1.6%   8.5%       27478     309.838750   3.00
+davit_large                 0 1     290.43   0.7%   5.6%       34016     290.434243   1.00
+davit_large-multi           0 1     290.77   0.7%   5.4%       34260     290.773492   5.00
+dlrm                        0 1  418230.56   0.1%   0.4%        7120  418230.564140   1.00
+focalnet                    0 1     381.31   0.5%   3.5%       25794     381.313064   2.00
+fp16                        0 1     252.71   0.1%   0.5%        1804     252.706435   0.00
+fp32                        0 1      18.93   0.0%   0.3%        2182      18.931403   0.00
+llama                       0 1     453.97  11.5%  71.6%       28442     453.966564   1.00
+reformer                    0 1      55.60   0.0%   0.2%       25420      55.599417   1.00
+regnet_y_128gf              0 1      78.11   0.9%   6.5%       31570      78.106848   2.00
+resnet152                   0 1     637.53   0.8%   5.9%       35958     637.529996   1.00
+resnet152-multi             0 1     638.26   0.7%   5.6%       35422     638.255791   5.00
+resnet50                    0 1    1013.93   1.6%  12.6%        4746    1013.930919   1.00
+rwkv                        1 1        NaN    NaN    NaN        1574            NaN   1.00
+stargan                     0 1      38.18   3.0%  22.9%       37442      38.177830   1.00
+super-slomo                 0 1      41.83   1.2%   9.6%       33816      41.828308   1.00
+t5                          0 1      46.26   0.6%   4.3%       35460      46.264485   2.00
+tf32                        0 1     133.34   0.0%   0.4%        2182     133.344138   0.00
+whisper                     0 1     214.83   0.1%   0.5%       36740     214.828130   1.00
+
+Scores
+------
+Failure rate:       3.70% (FAIL)
+Score:             245.06
+```

paice-v1-10-g1243bba/NVIDIA_A100_80GB_PCIe/majaguma.2024-04-02_16:55:21.895752/opt-1_3b-multinode.data

@@ -0,0 +1 @@
+{"event": "message", "data": {"message": "Skip opt-1_3b-multinode because the following capability is not satisfied: len(nodes) >= 2"}, "pipe": null}

paice-v1-10-g1243bba/NVIDIA_A100_80GB_PCIe/majaguma.2024-04-02_16:55:21.895752/opt-1_3b.data

@@ -0,0 +1 @@
+{"event": "error", "data": {"type": "KeyError", "message": "'port'", "trace": "Traceback (most recent call last):\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/multi.py\", line 202, in do_run\n    await exec_plan.execute(\"run\", timeout=True, timeout_delay=600)\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 108, in execute\n    return await execute_command(self, phase, timeout, timeout_delay, **kwargs)\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/executors.py\", line 57, in execute_command\n    for pack, argv, _kwargs in command.commands():\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 150, in commands\n    yield from executor.commands()\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 150, in commands\n    yield from executor.commands()\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 125, in commands\n    yield self.pack, self.argv(), self.kwargs()\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 121, in argv\n    return self._argv(**kwargs) + self.exec.argv(**kwargs)\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 304, in argv\n    script_args = self.exec.argv(**kwargs)\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 122, in argv\n    return self._argv(**kwargs)\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 682, in _argv\n    f\"--main_process_port={manager['port']}\",\nKeyError: 'port'\n"}, "pipe": null}

paice-v1-10-g1243bba/NVIDIA_A100_80GB_PCIe/majaguma.2024-04-02_16:55:21.895752/opt-6_7b-multinode.data

@@ -0,0 +1 @@
+{"event": "message", "data": {"message": "Skip opt-6_7b-multinode because the following capability is not satisfied: len(nodes) >= 2"}, "pipe": null}

paice-v1-10-g1243bba/NVIDIA_A100_80GB_PCIe/majaguma.2024-04-02_16:55:21.895752/opt-6_7b.data

@@ -0,0 +1 @@
+{"event": "error", "data": {"type": "KeyError", "message": "'port'", "trace": "Traceback (most recent call last):\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/multi.py\", line 202, in do_run\n    await exec_plan.execute(\"run\", timeout=True, timeout_delay=600)\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 108, in execute\n    return await execute_command(self, phase, timeout, timeout_delay, **kwargs)\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/executors.py\", line 57, in execute_command\n    for pack, argv, _kwargs in command.commands():\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 150, in commands\n    yield from executor.commands()\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 150, in commands\n    yield from executor.commands()\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 125, in commands\n    yield self.pack, self.argv(), self.kwargs()\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 121, in argv\n    return self._argv(**kwargs) + self.exec.argv(**kwargs)\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 304, in argv\n    script_args = self.exec.argv(**kwargs)\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 122, in argv\n    return self._argv(**kwargs)\n  File \"/Users/satyaortiz-gagne/travail/mila/CODE/milabench/milabench/commands/__init__.py\", line 682, in _argv\n    f\"--main_process_port={manager['port']}\",\nKeyError: 'port'\n"}, "pipe": null}