Allow TesseractSinterDecoder to be compiled with different DEMs #130

dandragona-dev · 2025-10-07T01:47:12Z

build_det_orders() is now called within compile_decoder_for_dem() to ensure the right configuration (for the DEM) is being used in each Sinter task. Previously, the TesseractConfig would always be using a single det_orders despite it depending upon the DEM for the task.

Feature request for TesseractSinterDecoder requires DEM to be configured correctly #129.
Refactored TesseractSinterDecoder constructor to have parameters which are then passed to a TessertConfig during compile_decoder_for_dem().
Added test to cover the crash referenced in issues/129.
Uses default values from TesseractConfig when applicable, otherwise improvises for num_det_orders, det_order_method, and seed.

oscarhiggott · 2025-10-21T03:02:31Z

Thank you, this looks excellent!

I wonder if it might worth including the "tesseract-long-beam" and "tesseract-short-beam" sinter decoders from my issue (#129 ) in the dictionary returned by tesseract_decoder.tesseract_sinter_compat.make_tesseract_sinter_decoders_dict?

oscarhiggott · 2025-10-21T03:04:58Z

We could also add a small example of how to use Tesseract with sinter in the main Tesseract README.md at the root of the directory (as is done in the chromobius repo, for example). This would make it much easier for new users to get started.

dandragona-dev · 2025-10-21T15:45:19Z

Sounds good, I'll add those suggestions in!

oscarhiggott · 2025-10-23T00:31:22Z

README.md

 ```
+## Using Tesseract with Sinter
+
+Tesseract can be easily integrated into [Sinter](https://github.com/quantumlib/Sinter) workflows. Sinter is a tool for running and organizing quantum error correction simulations. The `tesseract_sinter_compat` module provides the necessary interface.


Suggested change

Tesseract can be easily integrated into [Sinter](https://github.com/quantumlib/Sinter) workflows. Sinter is a tool for running and organizing quantum error correction simulations. The `tesseract_sinter_compat` module provides the necessary interface.

Tesseract can be easily integrated into [sinter](https://github.com/quantumlib/Stim/tree/main/glue/sample) workflows. Sinter is a tool for running and organizing quantum error correction simulations. The `tesseract_sinter_compat` module provides the necessary interface.

Good catch, thanks.

README.md

oscarhiggott · 2025-10-23T00:43:44Z

README.md

+#  Logical error rate: 0.0153
+```
+
+This example runs simulations for a repetition code with different distances [3, 5, 7] with different Tesseract default decoders. Sinter efficiently manages the execution of these tasks, and Tesseract is used for decoding. For more usage examples, see the tests in `src/py/tesseract_sinter_compat_test.py`.


Perhaps we could add a command line example here too, e.g.:

This example runs simulations for a repetition code with different distances [3, 5, 7] with different Tesseract default decoders.

Sinter can also be used at the command line. Here is an example of this using Tesseract:

sinter collect \ --circuits "example_circuit.stim" \ --decoders tesseract-long-beam \ --custom_decoders_module_function "tesseract_decoder:make_tesseract_sinter_decoders_dict" \ --max_shots 100_000 \ --max_errors 100 --processes auto \ --save_resume_filepath "stats.csv" \

Sinter efficiently manages the execution of these tasks, and Tesseract is used for decoding. For more usage examples, see the tests in src/py/tesseract_sinter_compat_test.py.

SG I'll add that in.

TesseractSinterDecoder visible from the root module.

README.md

oscarhiggott · 2025-10-24T05:04:22Z

README.md

+    tasks=tasks,
+    max_shots=10000,
+    decoders=decoders,
+    custom_decoders=tesseract_module.make_tesseract_sinter_decoders_dict(),


Suggested change

custom_decoders=tesseract_module.make_tesseract_sinter_decoders_dict(),

custom_decoders=make_tesseract_sinter_decoders_dict(),

oscarhiggott · 2025-10-24T05:22:50Z

README.md

+    decoders=decoders,
+    custom_decoders=tesseract_module.make_tesseract_sinter_decoders_dict(),
+)
+


I think it might be good to include one example of using the tesseract_decoder.TesseractSinterDecoder constructor with a full set of custom arguments. E.g. here's a full example (though maybe could be made more concise if combined with the example above):

import stim import sinter from tesseract_decoder import TesseractSinterDecoder import tesseract_decoder if __name__ == "__main__": custom_sinter_decoder = TesseractSinterDecoder( det_beam=10, beam_climbing=True, no_revisit_dets=True, merge_errors=True, pqlimit=1_000, num_det_orders=5, det_order_method=tesseract_decoder.utils.DetOrder.DetIndex, seed=2384753, ) p = 0.005 tasks = [ sinter.Task( circuit=stim.Circuit.generated( "surface_code:rotated_memory_x", distance=d, rounds=d, after_clifford_depolarization=p, ), json_metadata={"d": d, "r": d, "p": p}, ) for d in (3, 5) ] results = sinter.collect( num_workers=2, tasks=tasks, max_shots=10_000, decoders=["custom-tesseract-decoder"], custom_decoders={"custom-tesseract-decoder": custom_sinter_decoder}, print_progress=True, )

oscarhiggott · 2025-10-24T05:24:48Z

README.md

+        decoder=decoders[i],
+        json_metadata={"d": distance, "decoder": decoders[i]},
+    ))
+


Note that from here on needs to be after an:

if __name__ == "__main__":

Might be worth making sure that the example code works right away after copy-pasting (e.g. without the above fix the sinter error can be a bit confusing...)

Made some tweaks and tested it in a standalone python file.

bit of a refactor to make the code more concise. Tested it in a standalone python file as well.

oscarhiggott · 2025-10-24T20:31:02Z

README.md

+    for result in results:
+        print(f"task metadata = {result.json_metadata}")
+        print(f" Shots run: {result.shots}")
+        print(f" Observed errors: {result.errors}")
+        print(f" Logical error rate: {result.errors / result.shots}")
+
+    # Should get something like:
+    # task metadata = {'d': 3, 'decoder': 'tesseract'}
+    #  Shots run: 10000
+    #  Observed errors: 48
+    #  Logical error rate: 0.0048
+    # task metadata = {'d': 3, 'decoder': 'custom-tesseract-decoder'}
+    #  Shots run: 10000
+    #  Observed errors: 64
+    #  Logical error rate: 0.0064
+    # task metadata = {'d': 5, 'decoder': 'tesseract-short-beam'}
+    #  Shots run: 10000
+    #  Observed errors: 13
+    #  Logical error rate: 0.0013
+    # task metadata = {'d': 5, 'decoder': 'custom-tesseract-decoder'}
+    #  Shots run: 10000
+    #  Observed errors: 12
+    #  Logical error rate: 0.0012
+    # task metadata = {'d': 3, 'decoder': 'tesseract-long-beam'}
+    #  Shots run: 10000
+    #  Observed errors: 42
+    #  Logical error rate: 0.0042
+    # task metadata = {'d': 3, 'decoder': 'tesseract-short-beam'}
+    #  Shots run: 10000
+    #  Observed errors: 39
+    #  Logical error rate: 0.0039
+    # task metadata = {'d': 5, 'decoder': 'tesseract'}
+    #  Shots run: 10000
+    #  Observed errors: 15
+    #  Logical error rate: 0.0015
+    # task metadata = {'d': 5, 'decoder': 'tesseract-long-beam'}
+    #  Shots run: 10000
+    #  Observed errors: 13
+    #  Logical error rate: 0.0013
+    # task metadata = {'d': 7, 'decoder': 'tesseract'}
+    #  Shots run: 10000
+    #  Observed errors: 5
+    #  Logical error rate: 0.0005
+    # task metadata = {'d': 7, 'decoder': 'tesseract-long-beam'}
+    #  Shots run: 10000
+    #  Observed errors: 1
+    #  Logical error rate: 0.0001
+    # task metadata = {'d': 7, 'decoder': 'tesseract-short-beam'}
+    #  Shots run: 10000
+    #  Observed errors: 3
+    #  Logical error rate: 0.0003
+    # task metadata = {'d': 7, 'decoder': 'custom-tesseract-decoder'}
+    #  Shots run: 10000
+    #  Observed errors: 3
+    #  Logical error rate: 0.0003
+```


Looks great, just a nit: maybe we could just print out the csv instead as that will take up less vertical space on the README:

Suggested change

for result in results:

print(f"task metadata = {result.json_metadata}")

print(f" Shots run: {result.shots}")

print(f" Observed errors: {result.errors}")

print(f" Logical error rate: {result.errors / result.shots}")

# Should get something like:

# task metadata = {'d': 3, 'decoder': 'tesseract'}

# Shots run: 10000

# Observed errors: 48

# Logical error rate: 0.0048

# task metadata = {'d': 3, 'decoder': 'custom-tesseract-decoder'}

# Shots run: 10000

# Observed errors: 64

# Logical error rate: 0.0064

# task metadata = {'d': 5, 'decoder': 'tesseract-short-beam'}

# Shots run: 10000

# Observed errors: 13

# Logical error rate: 0.0013

# task metadata = {'d': 5, 'decoder': 'custom-tesseract-decoder'}

# Shots run: 10000

# Observed errors: 12

# Logical error rate: 0.0012

# task metadata = {'d': 3, 'decoder': 'tesseract-long-beam'}

# Shots run: 10000

# Observed errors: 42

# Logical error rate: 0.0042

# task metadata = {'d': 3, 'decoder': 'tesseract-short-beam'}

# Shots run: 10000

# Observed errors: 39

# Logical error rate: 0.0039

# task metadata = {'d': 5, 'decoder': 'tesseract'}

# Shots run: 10000

# Observed errors: 15

# Logical error rate: 0.0015

# task metadata = {'d': 5, 'decoder': 'tesseract-long-beam'}

# Shots run: 10000

# Observed errors: 13

# Logical error rate: 0.0013

# task metadata = {'d': 7, 'decoder': 'tesseract'}

# Shots run: 10000

# Observed errors: 5

# Logical error rate: 0.0005

# task metadata = {'d': 7, 'decoder': 'tesseract-long-beam'}

# Shots run: 10000

# Observed errors: 1

# Logical error rate: 0.0001

# task metadata = {'d': 7, 'decoder': 'tesseract-short-beam'}

# Shots run: 10000

# Observed errors: 3

# Logical error rate: 0.0003

# task metadata = {'d': 7, 'decoder': 'custom-tesseract-decoder'}

# Shots run: 10000

# Observed errors: 3

# Logical error rate: 0.0003

```

# Print samples as CSV data.

print(sinter.CSV_HEADER)

for sample in results:

print(sample.to_csv_line())

should get something like:

shots, errors, discards, seconds,decoder,strong_id,json_metadata,custom_counts 10000, 42, 0, 0.071,tesseract,1b3fce6286e438f38c00c8f6a5005947373515ab08e6446a7dd9ecdbef12d4cc,"{""d"":3,""decoder"":""tesseract""}", 10000, 49, 0, 0.546,custom-tesseract-decoder,7b082bec7541be858e239d7828a432e329cd448356bbdf051b8b8aa76c86625a,"{""d"":3,""decoder"":""custom-tesseract-decoder""}", 10000, 13, 0, 7.64,tesseract-long-beam,217a3542f56319924576658a6da7081ea2833f5167cf6d77fbc7071548e386a9,"{""d"":5,""decoder"":""tesseract-long-beam""}", 10000, 42, 0, 0.743,tesseract-short-beam,cf4a4b0ce0e4c7beec1171f58eddffe403ed7359db5016fca2e16174ea577057,"{""d"":3,""decoder"":""tesseract-short-beam""}", 10000, 34, 0, 0.924,tesseract-long-beam,8cfa0f2e4061629e13bc98fe213285dc00eb90f21bba36e08c76bcdf213a1c09,"{""d"":3,""decoder"":""tesseract-long-beam""}", 10000, 10, 0, 0.439,tesseract,8274ea5ffec15d6e71faed5ee1057cdd7e497cbaee4c6109784f8a74669d7f96,"{""d"":5,""decoder"":""tesseract""}", 10000, 8, 0, 3.93,custom-tesseract-decoder,8e4f5ab5dde00fec74127eea39ea52d5a98ae6ccfc277b5d9be450f78acc1c45,"{""d"":5,""decoder"":""custom-tesseract-decoder""}", 10000, 10, 0, 5.74,tesseract-short-beam,bf696535d62a25720c3a0c624ec5624002efe3f6cb0468963eee702efb48abc1,"{""d"":5,""decoder"":""tesseract-short-beam""}", 10000, 5, 0, 1.27,tesseract,3f94c61f1503844df6cf0d200b74ac01bfbc5e29e70cedbfc2faad67047e7887,"{""d"":7,""decoder"":""tesseract""}", 10000, 4, 0, 25.0,tesseract-long-beam,4d510f0acf511e24a833a93c956b683346696d8086866fadc73063fb09014c23,"{""d"":7,""decoder"":""tesseract-long-beam""}", 10000, 1, 0, 18.6,tesseract-short-beam,75782ce4593022fcedad4c73104711f05c9c635db92869531f78da336945b121,"{""d"":7,""decoder"":""tesseract-short-beam""}", 10000, 4, 0, 11.6,custom-tesseract-decoder,48f256a28fff47c58af7bffdf98fdee1d41a721751ee965c5d3c5712ac795dc8,"{""d"":7,""decoder"":""custom-tesseract-decoder""}",

dandragona-dev force-pushed the sinter-compat-integration branch 2 times, most recently from e8dc93e to a686e65 Compare October 7, 2025 15:15

oscarhiggott self-requested a review October 17, 2025 19:00

Update BUILD to remove "-march=native", for macOS builds

2168aa5

dandragona-dev force-pushed the sinter-compat-integration branch from a686e65 to 2168aa5 Compare October 23, 2025 00:29

oscarhiggott reviewed Oct 23, 2025

View reviewed changes

README.md Outdated Show resolved Hide resolved

oscarhiggott reviewed Oct 23, 2025

View reviewed changes

Added CLI example, and made make_tesseract_sinter_decoders_dic and

a1a42f3

TesseractSinterDecoder visible from the root module.

dandragona-dev force-pushed the sinter-compat-integration branch from 85b3230 to a1a42f3 Compare October 24, 2025 02:42

oscarhiggott reviewed Oct 24, 2025

View reviewed changes

README.md Outdated Show resolved Hide resolved

oscarhiggott reviewed Oct 24, 2025

View reviewed changes

README.md Outdated Show resolved Hide resolved

oscarhiggott reviewed Oct 24, 2025

View reviewed changes

Added a custom TesseractSinterDecoder to the python example and did a

18ab131

bit of a refactor to make the code more concise. Tested it in a standalone python file as well.

oscarhiggott reviewed Oct 24, 2025

View reviewed changes

Print a snippet of the CSV file after the Sinter collection.

9bd1108

	Tesseract can be easily integrated into [Sinter](https://github.com/quantumlib/Sinter) workflows. Sinter is a tool for running and organizing quantum error correction simulations. The `tesseract_sinter_compat` module provides the necessary interface.
	Tesseract can be easily integrated into [sinter](https://github.com/quantumlib/Stim/tree/main/glue/sample) workflows. Sinter is a tool for running and organizing quantum error correction simulations. The `tesseract_sinter_compat` module provides the necessary interface.

	custom_decoders=tesseract_module.make_tesseract_sinter_decoders_dict(),
	custom_decoders=make_tesseract_sinter_decoders_dict(),

Allow TesseractSinterDecoder to be compiled with different DEMs #130

Are you sure you want to change the base?

Allow TesseractSinterDecoder to be compiled with different DEMs #130

Conversation

dandragona-dev commented Oct 7, 2025

Uh oh!

oscarhiggott commented Oct 21, 2025

Uh oh!

oscarhiggott commented Oct 21, 2025

Uh oh!

dandragona-dev commented Oct 21, 2025

Uh oh!

oscarhiggott Oct 23, 2025

Choose a reason for hiding this comment

Uh oh!

dandragona-dev Oct 23, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

oscarhiggott Oct 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

dandragona-dev Oct 23, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

oscarhiggott Oct 24, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

oscarhiggott Oct 24, 2025

Choose a reason for hiding this comment

Uh oh!

oscarhiggott Oct 24, 2025

Choose a reason for hiding this comment

Uh oh!

dandragona-dev Oct 24, 2025

Choose a reason for hiding this comment

Uh oh!

oscarhiggott Oct 24, 2025

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

oscarhiggott Oct 23, 2025 •

edited

Loading

oscarhiggott Oct 24, 2025 •

edited

Loading