Last active
August 9, 2021 13:34
-
-
Save morganmcg1/0e4344df49fe3b43243505992ce998d5 to your computer and use it in GitHub Desktop.
gpt-j generation error stacktrace
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
2021-08-09 13:33:24.972717: E external/org_tensorflow/tensorflow/compiler/xla/pjrt/pjrt_stream_executor_client.cc:1981] Execution of replica 2 failed: Resource exhausted: Attempting to reserve 4.44G at the bottom of memory. That was not possible. There are 9.62G free, 0B reserved, and 2.65G reservable. | |
2021-08-09 13:33:24.972816: E external/org_tensorflow/tensorflow/compiler/xla/pjrt/pjrt_stream_executor_client.cc:1981] Execution of replica 5 failed: Resource exhausted: Attempting to reserve 4.44G at the bottom of memory. That was not possible. There are 9.62G free, 0B reserved, and 2.65G reservable. | |
2021-08-09 13:33:24.972875: E external/org_tensorflow/tensorflow/compiler/xla/pjrt/pjrt_stream_executor_client.cc:1981] Execution of replica 3 failed: Resource exhausted: Attempting to reserve 4.44G at the bottom of memory. That was not possible. There are 9.62G free, 0B reserved, and 2.65G reservable. | |
2021-08-09 13:33:24.972954: E external/org_tensorflow/tensorflow/compiler/xla/pjrt/pjrt_stream_executor_client.cc:1981] Execution of replica 1 failed: Resource exhausted: Attempting to reserve 4.44G at the bottom of memory. That was not possible. There are 9.62G free, 0B reserved, and 2.65G reservable. | |
2021-08-09 13:33:24.972999: E external/org_tensorflow/tensorflow/compiler/xla/pjrt/pjrt_stream_executor_client.cc:1981] Execution of replica 7 failed: Resource exhausted: Attempting to reserve 4.44G at the bottom of memory. That was not possible. There are 9.62G free, 0B reserved, and 2.65G reservable. | |
2021-08-09 13:33:24.973251: E external/org_tensorflow/tensorflow/compiler/xla/pjrt/pjrt_stream_executor_client.cc:1981] Execution of replica 6 failed: Resource exhausted: Attempting to reserve 4.44G at the bottom of memory. That was not possible. There are 9.62G free, 0B reserved, and 2.65G reservable. | |
2021-08-09 13:33:24.974006: E external/org_tensorflow/tensorflow/compiler/xla/pjrt/pjrt_stream_executor_client.cc:1981] Execution of replica 4 failed: Resource exhausted: Attempting to reserve 4.44G at the bottom of memory. That was not possible. There are 9.62G free, 0B reserved, and 2.65G reservable. | |
2021-08-09 13:33:25.109534: E external/org_tensorflow/tensorflow/compiler/xla/pjrt/pjrt_stream_executor_client.cc:1981] Execution of replica 0 failed: Resource exhausted: Attempting to reserve 4.44G at the bottom of memory. That was not possible. There are 9.61G free, 0B reserved, and 2.65G reservable. | |
Traceback (most recent call last): | |
File "device_train.py", line 384, in <module> | |
loss, last_loss, grad_norm, grad_norm_micro = train_step( | |
File "device_train.py", line 113, in train_step | |
loss, last_loss, grad_norm, grad_norm_micro = network.train(inputs) | |
File "/home/morganmcguire/mesh-transformer-jax/mesh_transformer/transformer_shard.py", line 301, in train | |
loss, last_loss, grad_norm, grad_norm_micro, self.state = self.train_xmap(self.state, obs, target) | |
File "/usr/local/lib/python3.8/dist-packages/jax/experimental/maps.py", line 516, in fun_mapped | |
out_flat = xmap_p.bind( | |
File "/usr/local/lib/python3.8/dist-packages/jax/experimental/maps.py", line 652, in bind | |
return core.call_bind(self, fun, *args, **params) # type: ignore | |
File "/usr/local/lib/python3.8/dist-packages/jax/core.py", line 1393, in call_bind | |
outs = primitive.process(top_trace, fun, tracers, params) | |
File "/usr/local/lib/python3.8/dist-packages/jax/experimental/maps.py", line 655, in process | |
return trace.process_xmap(self, fun, tracers, params) | |
File "/usr/local/lib/python3.8/dist-packages/jax/core.py", line 600, in process_call | |
return primitive.impl(f, *tracers, **params) | |
File "/usr/local/lib/python3.8/dist-packages/jax/experimental/maps.py", line 539, in xmap_impl | |
return make_xmap_callable(fun, name, in_axes, out_axes_thunk, donated_invars, global_axis_sizes, | |
File "/usr/local/lib/python3.8/dist-packages/jax/interpreters/pxla.py", line 1130, in execute_replicated | |
out_bufs = compiled.execute_sharded_on_local_devices(input_bufs) | |
RuntimeError: Resource exhausted: Attempting to reserve 4.44G at the bottom of memory. That was not possible. There are 9.61G free, 0B reserved, and 2.65G reservable.: while running replica 0 and partition 0 of a replicated computation (other replicas may have failed as well). | |
Traceback (most recent call last): | |
File "device_train.py", line 384, in <module> | |
loss, last_loss, grad_norm, grad_norm_micro = train_step( | |
File "device_train.py", line 113, in train_step | |
loss, last_loss, grad_norm, grad_norm_micro = network.train(inputs) | |
File "/home/morganmcguire/mesh-transformer-jax/mesh_transformer/transformer_shard.py", line 301, in train | |
loss, last_loss, grad_norm, grad_norm_micro, self.state = self.train_xmap(self.state, obs, target) | |
File "/usr/local/lib/python3.8/dist-packages/jax/experimental/maps.py", line 516, in fun_mapped | |
out_flat = xmap_p.bind( | |
File "/usr/local/lib/python3.8/dist-packages/jax/experimental/maps.py", line 652, in bind | |
return core.call_bind(self, fun, *args, **params) # type: ignore | |
File "/usr/local/lib/python3.8/dist-packages/jax/core.py", line 1393, in call_bind | |
outs = primitive.process(top_trace, fun, tracers, params) | |
File "/usr/local/lib/python3.8/dist-packages/jax/experimental/maps.py", line 655, in process | |
return trace.process_xmap(self, fun, tracers, params) | |
File "/usr/local/lib/python3.8/dist-packages/jax/core.py", line 600, in process_call | |
return primitive.impl(f, *tracers, **params) | |
File "/usr/local/lib/python3.8/dist-packages/jax/experimental/maps.py", line 539, in xmap_impl | |
return make_xmap_callable(fun, name, in_axes, out_axes_thunk, donated_invars, global_axis_sizes, | |
File "/usr/local/lib/python3.8/dist-packages/jax/interpreters/pxla.py", line 1130, in execute_replicated | |
out_bufs = compiled.execute_sharded_on_local_devices(input_bufs) | |
RuntimeError: Resource exhausted: Attempting to reserve 4.44G at the bottom of memory. That was not possible. There are 9.61G free, 0B reserved, and 2.65G reservable.: while running replica 0 and partition 0 of a replicated computation (other replicas may have failed as well). |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment