-
Notifications
You must be signed in to change notification settings - Fork 104
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
1 parent
941b34c
commit ebb919e
Showing
12 changed files
with
1,807 additions
and
0 deletions.
There are no files selected for viewing
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
178 changes: 178 additions & 0 deletions
178
tests/sweep_framework/sweeps/eltwise/binary/eq/eq_survey.py
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,178 @@ | ||
| # SPDX-FileCopyrightText: © 2024 Tenstorrent Inc. | ||
|
|
||
| # SPDX-License-Identifier: Apache-2.0 | ||
|
|
||
| from typing import Optional, Tuple | ||
| from functools import partial | ||
|
|
||
| import torch | ||
| import ttnn | ||
| from tests.tt_eager.python_api_testing.sweep_tests.generation_funcs import gen_func_with_cast_tt | ||
|
|
||
| from tests.ttnn.utils_for_testing import check_with_pcc, start_measuring_time, stop_measuring_time | ||
| from models.utility_functions import torch_random | ||
|
|
||
| # Parameters provided to the test vector generator are defined here. | ||
| # They are defined as dict-type suites that contain the arguments to the run function as keys, and lists of possible inputs as values. | ||
| # Each suite has a key name (in this case "suite_1") which will associate the test vectors to this specific suite of inputs. | ||
| # Developers can create their own generator functions and pass them to the parameters as inputs. | ||
| parameters = { | ||
| "lte_survey-4": { | ||
| "input_shape": [{"self": [1, 1, 512, 512], "other": [1, 1, 512, 512]}], | ||
| "input_a_dtype": [ttnn.bfloat4_b], | ||
| "input_b_dtype": [ttnn.bfloat4_b], | ||
| "input_a_layout": [ttnn.TILE_LAYOUT], | ||
| "input_b_layout": [ttnn.TILE_LAYOUT], | ||
| "input_memory_config": [ | ||
| {"a_mem_config": "l1_interleaved", "b_mem_config": "l1_interleaved"}, | ||
| {"a_mem_config": "dram_interleaved", "b_mem_config": "l1_interleaved"}, | ||
| {"a_mem_config": "l1_interleaved", "b_mem_config": "dram_interleaved"}, | ||
| {"a_mem_config": "dram_interleaved", "b_mem_config": "dram_interleaved"}, | ||
| {"a_mem_config": "dram_interleaved", "b_mem_config": "l1_height_sharded_rm"}, | ||
| {"a_mem_config": "dram_interleaved", "b_mem_config": "l1_width_sharded_rm"}, | ||
| {"a_mem_config": "dram_interleaved", "b_mem_config": "l1_block_sharded_rm"}, | ||
| {"a_mem_config": "dram_interleaved", "b_mem_config": "l1_height_sharded_cm"}, | ||
| {"a_mem_config": "dram_interleaved", "b_mem_config": "l1_width_sharded_cm"}, | ||
| {"a_mem_config": "dram_interleaved", "b_mem_config": "l1_block_sharded_cm"}, | ||
| {"a_mem_config": "l1_height_sharded_rm", "b_mem_config": "dram_interleaved"}, | ||
| {"a_mem_config": "l1_width_sharded_rm", "b_mem_config": "dram_interleaved"}, | ||
| {"a_mem_config": "l1_block_sharded_rm", "b_mem_config": "dram_interleaved"}, | ||
| {"a_mem_config": "l1_height_sharded_cm", "b_mem_config": "dram_interleaved"}, | ||
| {"a_mem_config": "l1_width_sharded_cm", "b_mem_config": "dram_interleaved"}, | ||
| {"a_mem_config": "l1_block_sharded_cm", "b_mem_config": "dram_interleaved"}, | ||
| {"a_mem_config": "l1_height_sharded_rm", "b_mem_config": "l1_height_sharded_rm"}, | ||
| {"a_mem_config": "l1_height_sharded_cm", "b_mem_config": "l1_height_sharded_cm"}, | ||
| {"a_mem_config": "l1_width_sharded_rm", "b_mem_config": "l1_width_sharded_rm"}, | ||
| {"a_mem_config": "l1_width_sharded_cm", "b_mem_config": "l1_width_sharded_cm"}, | ||
| {"a_mem_config": "l1_block_sharded_rm", "b_mem_config": "l1_block_sharded_rm"}, | ||
| {"a_mem_config": "l1_block_sharded_cm", "b_mem_config": "l1_block_sharded_cm"}, | ||
| ], | ||
| }, | ||
| } | ||
|
|
||
|
|
||
| def return_mem_config(mem_config_string): | ||
| if mem_config_string == "l1_interleaved": | ||
| return ttnn.L1_MEMORY_CONFIG | ||
| elif mem_config_string == "dram_interleaved": | ||
| return ttnn.DRAM_MEMORY_CONFIG | ||
| elif mem_config_string == "l1_height_sharded_rm": | ||
| return ttnn.create_sharded_memory_config( | ||
| shape=(512 // 8, 512), | ||
| core_grid=ttnn.CoreGrid(y=2, x=4), | ||
| strategy=ttnn.ShardStrategy.HEIGHT, | ||
| orientation=ttnn.ShardOrientation.ROW_MAJOR, | ||
| use_height_and_width_as_shard_shape=True, | ||
| ) | ||
| elif mem_config_string == "l1_width_sharded_rm": | ||
| return ttnn.create_sharded_memory_config( | ||
| shape=(512, 512 // 8), | ||
| core_grid=ttnn.CoreGrid(y=2, x=4), | ||
| strategy=ttnn.ShardStrategy.WIDTH, | ||
| orientation=ttnn.ShardOrientation.ROW_MAJOR, | ||
| use_height_and_width_as_shard_shape=True, | ||
| ) | ||
| elif mem_config_string == "l1_block_sharded_rm": | ||
| return ttnn.create_sharded_memory_config( | ||
| shape=(512 // 2, 512 // 4), | ||
| core_grid=ttnn.CoreGrid(y=2, x=4), | ||
| strategy=ttnn.ShardStrategy.BLOCK, | ||
| orientation=ttnn.ShardOrientation.ROW_MAJOR, | ||
| use_height_and_width_as_shard_shape=True, | ||
| ) | ||
| elif mem_config_string == "l1_height_sharded_cm": | ||
| return ttnn.create_sharded_memory_config( | ||
| shape=(512, 512 // 8), | ||
| core_grid=ttnn.CoreGrid(y=2, x=4), | ||
| strategy=ttnn.ShardStrategy.HEIGHT, | ||
| orientation=ttnn.ShardOrientation.COL_MAJOR, | ||
| use_height_and_width_as_shard_shape=True, | ||
| ) | ||
| elif mem_config_string == "l1_width_sharded_cm": | ||
| return ttnn.create_sharded_memory_config( | ||
| shape=(512 // 8, 512), | ||
| core_grid=ttnn.CoreGrid(y=2, x=4), | ||
| strategy=ttnn.ShardStrategy.WIDTH, | ||
| orientation=ttnn.ShardOrientation.COL_MAJOR, | ||
| use_height_and_width_as_shard_shape=True, | ||
| ) | ||
| elif mem_config_string == "l1_block_sharded_cm": | ||
| return ttnn.create_sharded_memory_config( | ||
| shape=(512 // 2, 512 // 4), | ||
| core_grid=ttnn.CoreGrid(y=2, x=4), | ||
| strategy=ttnn.ShardStrategy.BLOCK, | ||
| orientation=ttnn.ShardOrientation.COL_MAJOR, | ||
| use_height_and_width_as_shard_shape=True, | ||
| ) | ||
| raise ("Input mem_config_string is not valid!") | ||
|
|
||
|
|
||
| # This is the run instructions for the test, defined by the developer. | ||
| # The run function must take the above-defined parameters as inputs. | ||
| # The runner will call this run function with each test vector, and the returned results from this function will be stored. | ||
| # If you defined a device_mesh_fixture above, the object you yielded will be passed into this function as 'device'. Otherwise, it will be the default ttnn device opened by the infra. | ||
| def run( | ||
| input_shape, | ||
| input_a_dtype, | ||
| input_b_dtype, | ||
| input_a_layout, | ||
| input_b_layout, | ||
| input_memory_config, | ||
| *, | ||
| device, | ||
| ) -> list: | ||
| torch.manual_seed(0) | ||
|
|
||
| input_a_memory_config = input_memory_config["a_mem_config"] | ||
| input_b_memory_config = input_memory_config["b_mem_config"] | ||
|
|
||
| torch_input_tensor_a = gen_func_with_cast_tt( | ||
| partial(torch_random, low=-100, high=100, dtype=torch.float32), input_a_dtype | ||
| )(input_shape["self"]) | ||
|
|
||
| if isinstance(input_shape["other"], list): | ||
| torch_input_tensor_b = gen_func_with_cast_tt( | ||
| partial(torch_random, low=-105, high=105, dtype=torch.float32), input_b_dtype | ||
| )(input_shape["other"]) | ||
| else: | ||
| torch_input_tensor_b = torch.tensor(input_shape["other"], dtype=torch.float32) | ||
|
|
||
| input_tensor_a = ttnn.from_torch( | ||
| torch_input_tensor_a, | ||
| dtype=input_a_dtype, | ||
| layout=input_a_layout, | ||
| device=device, | ||
| memory_config=return_mem_config(input_a_memory_config), | ||
| ) | ||
|
|
||
| input_tensor_b = ttnn.from_torch( | ||
| torch_input_tensor_b, | ||
| dtype=input_b_dtype, | ||
| layout=input_b_layout, | ||
| device=device, | ||
| memory_config=return_mem_config(input_b_memory_config), | ||
| ) | ||
|
|
||
| if input_a_dtype == ttnn.bfloat8_b: | ||
| torch_input_tensor_a = ttnn.to_torch(input_tensor_a) | ||
|
|
||
| if input_b_dtype == ttnn.bfloat8_b: | ||
| torch_input_tensor_b = ttnn.to_torch(input_tensor_b) | ||
|
|
||
| if input_a_dtype == ttnn.bfloat4_b: | ||
| torch_input_tensor_a = ttnn.to_torch(input_tensor_a) | ||
|
|
||
| if input_b_dtype == ttnn.bfloat4_b: | ||
| torch_input_tensor_b = ttnn.to_torch(input_tensor_b) | ||
|
|
||
| golden_function = ttnn.get_golden_function(ttnn.le) | ||
| torch_output_tensor = golden_function(torch_input_tensor_a, torch_input_tensor_b) | ||
| start_time = start_measuring_time() | ||
| result = ttnn.experimental.lte(input_tensor_a, input_tensor_b) | ||
| output_tensor = ttnn.to_torch(result) | ||
| torch.set_printoptions(linewidth=400, threshold=10000, precision=5, sci_mode=False, edgeitems=17) | ||
| print("torch_output_tensor", torch_output_tensor) | ||
| print("output_tensor", output_tensor) | ||
| e2e_perf = stop_measuring_time(start_time) | ||
|
|
||
| return [check_with_pcc(torch_output_tensor, output_tensor, pcc=0.999), e2e_perf] |
Oops, something went wrong.