Pull pyston benchmarks into pyperformance. #60

ericsnowcurrently · 2021-06-16T17:20:11Z

The Pyston project published a set of macro benchmarks that they used. We can work on incorporating those into the pyperformance suite.

TODO:

Steps to incorporate Pyston benchmarks on a benchmarking host:

clone the Pyston benchmarks repo
create a manifest file with just an "includes" section, containing references to the manifests you're combining
add a "manifest=" entry to the config file you're passing to pyperformance compile ...
proceed as usual

More concretely:

#git clone --branch benchmark-management https://github.com/ericsnowcurrently/python-performance pyperformance
#git clone --branch pyperformance https://github.com/ericsnowcurrently/pyston-macrobenchmarks
git clone https://github.com/pyston/python-macrobenchmarks pyston-macrobenchmarks

cat > MANIFEST.combined << EOF
[includes]
<default>
pyston-macrobenchmarks/benchmarks/MANIFEST
EOF

cat > ./cpython-perf.ini << EOF
[config]
...

[scm]
...

[compile]
...

[run_benchmark]
manifest = MANIFEST.combined
...
EOF

PYTHONPATH=./pyperformance python3 -m pyperformance compile ./cpython-perf.ini main

The text was updated successfully, but these errors were encountered:

Fidget-Spinner · 2021-06-18T15:29:49Z

FWIW, I started looking at the Pyston suite and tried to find areas where pyperformance lack a month ago. So far I ported over their mypy benchmark (see python/pyperformance#102).

Just my two cents: I think their aiohttp, gevent and gunicorn benchmarks are interesting too because they're more modern workloads. However, I've been trying to find benchmarks that use OOP concepts more heavily since that's an area pyperformance lacks right now.

ericsnowcurrently · 2021-07-23T20:22:29Z

I've mostly gotten to the point I can post PRs:

Now I just need to follow up with @vstinner and @pablogsal on the pyperformance approach.

ericsnowcurrently · 2021-09-01T16:30:02Z

Also, with my changes we could get rid of the idea of "groups" in pyperformance. Instead we would have a separate manifest file for each of the existing groups.

ericsnowcurrently · 2021-11-09T21:05:41Z

Current pyperformance vs. my branch:

$ python3.8 -m pyperf compare_to ~benchmarking/BENCH/REQUESTS/req-1636475547-esnow/results-data.json.gz ~benchmarking/BENCH/REQUESTS/req-1636488560-esnow/results-data.json.gz -G --table
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| Benchmark               | BENCH/REQUESTS/req-1636475547-esnow/results-data.json.gz | BENCH/REQUESTS/req-1636488560-esnow/results-data.json.gz |
+=========================+==========================================================+==========================================================+
| unpack_sequence         | 46.8 ns                                                  | 44.1 ns: 1.06x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| telco                   | 6.57 ms                                                  | 6.31 ms: 1.04x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| regex_effbot            | 3.22 ms                                                  | 3.12 ms: 1.03x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| float                   | 84.1 ms                                                  | 81.6 ms: 1.03x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| hexiom                  | 7.55 ms                                                  | 7.33 ms: 1.03x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| meteor_contest          | 108 ms                                                   | 105 ms: 1.03x faster                                     |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| logging_silent          | 118 ns                                                   | 114 ns: 1.03x faster                                     |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| pathlib                 | 19.4 ms                                                  | 19.0 ms: 1.02x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| nbody                   | 114 ms                                                   | 111 ms: 1.02x faster                                     |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| scimark_monte_carlo     | 80.5 ms                                                  | 78.8 ms: 1.02x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| xml_etree_parse         | 155 ms                                                   | 152 ms: 1.02x faster                                     |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| xml_etree_process       | 58.9 ms                                                  | 57.9 ms: 1.02x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| python_startup          | 8.51 ms                                                  | 8.37 ms: 1.02x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| json_dumps              | 12.5 ms                                                  | 12.3 ms: 1.02x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| json_loads              | 25.6 us                                                  | 25.2 us: 1.02x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| scimark_sor             | 155 ms                                                   | 153 ms: 1.01x faster                                     |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| raytrace                | 331 ms                                                   | 328 ms: 1.01x faster                                     |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| sqlite_synth            | 2.78 us                                                  | 2.75 us: 1.01x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| chaos                   | 77.4 ms                                                  | 76.5 ms: 1.01x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| regex_v8                | 23.2 ms                                                  | 23.0 ms: 1.01x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| xml_etree_generate      | 80.7 ms                                                  | 80.1 ms: 1.01x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| pickle_list             | 4.62 us                                                  | 4.59 us: 1.01x faster                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| xml_etree_iterparse     | 106 ms                                                   | 105 ms: 1.01x faster                                     |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| regex_compile           | 143 ms                                                   | 142 ms: 1.00x faster                                     |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| pidigits                | 186 ms                                                   | 187 ms: 1.00x slower                                     |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| regex_dna               | 214 ms                                                   | 215 ms: 1.00x slower                                     |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| python_startup_no_site  | 5.90 ms                                                  | 5.92 ms: 1.00x slower                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| unpickle_pure_python    | 266 us                                                   | 267 us: 1.00x slower                                     |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| logging_format          | 6.47 us                                                  | 6.52 us: 1.01x slower                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| pyflate                 | 529 ms                                                   | 535 ms: 1.01x slower                                     |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| scimark_fft             | 340 ms                                                   | 344 ms: 1.01x slower                                     |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| nqueens                 | 88.5 ms                                                  | 90.6 ms: 1.02x slower                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| pickle                  | 9.79 us                                                  | 10.1 us: 1.03x slower                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| unpickle_list           | 5.10 us                                                  | 5.27 us: 1.03x slower                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| pickle_dict             | 27.3 us                                                  | 28.2 us: 1.03x slower                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| richards                | 54.2 ms                                                  | 56.1 ms: 1.04x slower                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| scimark_sparse_mat_mult | 4.55 ms                                                  | 4.81 ms: 1.06x slower                                    |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+
| Geometric mean          | (ref)                                                    | 1.00x faster                                             |
+-------------------------+----------------------------------------------------------+----------------------------------------------------------+

Benchmark hidden because not significant (9): fannkuch, logging_simple, deltablue, go, spectral_norm, 2to3, pickle_pure_python, unpickle, scimark_lu
Ignored benchmarks (12) of BENCH/REQUESTS/req-1636475547-esnow/results-data.json.gz: chameleon, crypto_pyaes, django_template, dulwich_log, mako, sqlalchemy_declarative, sqlalchemy_imperative, sympy_expand, sympy_integrate, sympy_str, sympy_sum, tornado_http

Although the geometric means indicates no change, there is some variance in the results where I'd expect none. I'll be looking into that.

@brandtbucher, any ideas?

gvanrossum · 2021-11-09T21:15:27Z

What exactly do the two columns represent?

There is often some noise in benchmarks -- Brandt has tried to quantify this in #109

brandtbucher · 2021-11-09T21:23:11Z

Yeah, this is a problem with certain benchmarks. For example, I've often seen pidigits swing +/-11% on two builds of the same exact changeset. The two runs each showed no deviation individually:

pidigits: 188 ms +- 0 ms -> 208 ms +- 0 ms: 1.11x slower

brandtbucher · 2021-11-09T21:37:05Z

Also, I've grown to dislike pyperf's pretty "table" view, because it (for some reason) hides the standard deviation for each run. The "default" view often reveals overlapping ranges in the results. I'm no statistician, but I'm pretty sure that that hurts their statistical significance.

As a concrete example, here's the whole run that I got that pidigits result from. Notice that pidigits, pickle, and regex_effbot are the only "significant" results without overlapping error bars (and, as I've shown in #109, even those cannot be reliably trusted):

$ pyperf compare_to ../results/old/0/main.json.gz ../results/old/0/main-again.json.gz --group
Slower (27):
- pidigits: 188 ms +- 0 ms -> 208 ms +- 0 ms: 1.11x slower
- unpickle: 13.4 us +- 0.1 us -> 14.0 us +- 0.8 us: 1.04x slower
- unpickle_list: 4.98 us +- 0.06 us -> 5.12 us +- 0.10 us: 1.03x slower
- logging_format: 6.82 us +- 0.08 us -> 7.00 us +- 0.12 us: 1.03x slower
- pickle_list: 4.41 us +- 0.06 us -> 4.53 us +- 0.04 us: 1.03x slower
- json_loads: 25.3 us +- 0.7 us -> 25.9 us +- 0.3 us: 1.02x slower
- meteor_contest: 107 ms +- 1 ms -> 109 ms +- 4 ms: 1.02x slower
- scimark_monte_carlo: 78.7 ms +- 0.9 ms -> 80.1 ms +- 1.8 ms: 1.02x slower
- raytrace: 326 ms +- 2 ms -> 331 ms +- 3 ms: 1.02x slower
- scimark_sor: 156 ms +- 2 ms -> 158 ms +- 4 ms: 1.01x slower
- pickle: 9.87 us +- 0.09 us -> 10.0 us +- 0.1 us: 1.01x slower
- fannkuch: 421 ms +- 2 ms -> 426 ms +- 5 ms: 1.01x slower
- mako: 12.7 ms +- 0.1 ms -> 12.8 ms +- 0.1 ms: 1.01x slower
- xml_etree_generate: 83.9 ms +- 1.1 ms -> 84.7 ms +- 0.9 ms: 1.01x slower
- logging_simple: 6.10 us +- 0.08 us -> 6.16 us +- 0.09 us: 1.01x slower
- pyflate: 531 ms +- 4 ms -> 536 ms +- 6 ms: 1.01x slower
- logging_silent: 115 ns +- 2 ns -> 116 ns +- 1 ns: 1.01x slower
- xml_etree_parse: 154 ms +- 3 ms -> 156 ms +- 3 ms: 1.01x slower
- nbody: 113 ms +- 1 ms -> 114 ms +- 2 ms: 1.01x slower
- scimark_sparse_mat_mult: 4.90 ms +- 0.07 ms -> 4.94 ms +- 0.02 ms: 1.01x slower
- xml_etree_process: 61.7 ms +- 0.6 ms -> 62.2 ms +- 0.5 ms: 1.01x slower
- json_dumps: 12.5 ms +- 0.2 ms -> 12.6 ms +- 0.1 ms: 1.01x slower
- pickle_dict: 27.2 us +- 0.2 us -> 27.3 us +- 0.2 us: 1.01x slower
- regex_compile: 151 ms +- 1 ms -> 151 ms +- 1 ms: 1.01x slower
- django_template: 39.0 ms +- 0.4 ms -> 39.2 ms +- 0.3 ms: 1.01x slower
- go: 171 ms +- 2 ms -> 172 ms +- 2 ms: 1.01x slower
- dulwich_log: 68.6 ms +- 0.4 ms -> 68.8 ms +- 0.5 ms: 1.00x slower

Faster (17):
- regex_effbot: 3.31 ms +- 0.05 ms -> 3.11 ms +- 0.08 ms: 1.06x faster
- unpack_sequence: 47.6 ns +- 1.8 ns -> 46.1 ns +- 0.5 ns: 1.03x faster
- spectral_norm: 114 ms +- 3 ms -> 111 ms +- 2 ms: 1.03x faster
- nqueens: 89.6 ms +- 1.3 ms -> 88.3 ms +- 0.6 ms: 1.01x faster
- scimark_fft: 350 ms +- 3 ms -> 345 ms +- 3 ms: 1.01x faster
- regex_dna: 214 ms +- 1 ms -> 212 ms +- 2 ms: 1.01x faster
- sympy_integrate: 22.4 ms +- 0.1 ms -> 22.2 ms +- 0.1 ms: 1.01x faster
- regex_v8: 24.3 ms +- 0.1 ms -> 24.1 ms +- 0.5 ms: 1.01x faster
- chaos: 79.0 ms +- 1.0 ms -> 78.3 ms +- 0.8 ms: 1.01x faster
- tornado_http: 111 ms +- 3 ms -> 110 ms +- 3 ms: 1.01x faster
- unpickle_pure_python: 276 us +- 2 us -> 274 us +- 2 us: 1.01x faster
- pickle_pure_python: 379 us +- 3 us -> 376 us +- 3 us: 1.01x faster
- crypto_pyaes: 88.4 ms +- 0.5 ms -> 87.9 ms +- 1.0 ms: 1.01x faster
- sympy_expand: 515 ms +- 4 ms -> 512 ms +- 6 ms: 1.01x faster
- sympy_str: 307 ms +- 4 ms -> 306 ms +- 2 ms: 1.01x faster
- python_startup_no_site: 5.85 ms +- 0.01 ms -> 5.84 ms +- 0.00 ms: 1.00x faster
- python_startup: 8.46 ms +- 0.01 ms -> 8.45 ms +- 0.01 ms: 1.00x faster

Benchmark hidden because not significant (14): 2to3, chameleon, deltablue, float, hexiom, pathlib, richards, scimark_lu, sqlalchemy_declarative, sqlalchemy_imperative, sqlite_synth, sympy_sum, telco, xml_etree_iterparse

Geometric mean: 1.00x slower

ericsnowcurrently · 2021-11-09T21:53:29Z

FYI, the pyston benchmarks don't work against CPython main (as of about a month ago) due to dependencies failing to build. It looks like a problem in Cython. See cython/cython#4461 and MagicStack/uvloop#450.

gvanrossum · 2021-11-09T22:01:13Z

I wonder if we just need to have a low-intensity workstream that fixes various issues we're causing in Cython and other packages -- not just so 3rd parties can test with the alpha releases but so *we* can test with 3rd party packages.

ericsnowcurrently · 2021-11-09T22:01:25Z

What exactly do the two columns represent?

It's the two results files that I was comparing.

Here's the non-table comparison from pyperf:

Slower (13):
- scimark_sparse_mat_mult: 4.55 ms +- 0.05 ms -> 4.81 ms +- 0.08 ms: 1.06x slower
- richards: 54.2 ms +- 0.6 ms -> 56.1 ms +- 1.0 ms: 1.04x slower
- pickle_dict: 27.3 us +- 0.2 us -> 28.2 us +- 0.2 us: 1.03x slower
- unpickle_list: 5.10 us +- 0.05 us -> 5.27 us +- 0.09 us: 1.03x slower
- pickle: 9.79 us +- 0.07 us -> 10.1 us +- 0.2 us: 1.03x slower
- nqueens: 88.5 ms +- 1.0 ms -> 90.6 ms +- 0.8 ms: 1.02x slower
- scimark_fft: 340 ms +- 2 ms -> 344 ms +- 8 ms: 1.01x slower
- pyflate: 529 ms +- 4 ms -> 535 ms +- 8 ms: 1.01x slower
- logging_format: 6.47 us +- 0.09 us -> 6.52 us +- 0.10 us: 1.01x slower
- unpickle_pure_python: 266 us +- 2 us -> 267 us +- 2 us: 1.00x slower
- python_startup_no_site: 5.90 ms +- 0.00 ms -> 5.92 ms +- 0.00 ms: 1.00x slower
- regex_dna: 214 ms +- 1 ms -> 215 ms +- 0 ms: 1.00x slower
- pidigits: 186 ms +- 0 ms -> 187 ms +- 0 ms: 1.00x slower

Faster (24):
- unpack_sequence: 46.8 ns +- 1.2 ns -> 44.1 ns +- 0.8 ns: 1.06x faster
- telco: 6.57 ms +- 0.08 ms -> 6.31 ms +- 0.10 ms: 1.04x faster
- regex_effbot: 3.22 ms +- 0.01 ms -> 3.12 ms +- 0.06 ms: 1.03x faster
- float: 84.1 ms +- 0.9 ms -> 81.6 ms +- 1.0 ms: 1.03x faster
- hexiom: 7.55 ms +- 0.08 ms -> 7.33 ms +- 0.05 ms: 1.03x faster
- meteor_contest: 108 ms +- 3 ms -> 105 ms +- 2 ms: 1.03x faster
- logging_silent: 118 ns +- 2 ns -> 114 ns +- 2 ns: 1.03x faster
- pathlib: 19.4 ms +- 0.3 ms -> 19.0 ms +- 0.3 ms: 1.02x faster
- nbody: 114 ms +- 2 ms -> 111 ms +- 1 ms: 1.02x faster
- scimark_monte_carlo: 80.5 ms +- 1.0 ms -> 78.8 ms +- 2.0 ms: 1.02x faster
- xml_etree_parse: 155 ms +- 2 ms -> 152 ms +- 2 ms: 1.02x faster
- xml_etree_process: 58.9 ms +- 0.5 ms -> 57.9 ms +- 0.6 ms: 1.02x faster
- python_startup: 8.51 ms +- 0.01 ms -> 8.37 ms +- 0.01 ms: 1.02x faster
- json_dumps: 12.5 ms +- 0.1 ms -> 12.3 ms +- 0.1 ms: 1.02x faster
- json_loads: 25.6 us +- 0.3 us -> 25.2 us +- 0.2 us: 1.02x faster
- scimark_sor: 155 ms +- 3 ms -> 153 ms +- 1 ms: 1.01x faster
- raytrace: 331 ms +- 3 ms -> 328 ms +- 7 ms: 1.01x faster
- sqlite_synth: 2.78 us +- 0.07 us -> 2.75 us +- 0.06 us: 1.01x faster
- chaos: 77.4 ms +- 0.7 ms -> 76.5 ms +- 0.6 ms: 1.01x faster
- regex_v8: 23.2 ms +- 0.2 ms -> 23.0 ms +- 0.1 ms: 1.01x faster
- xml_etree_generate: 80.7 ms +- 0.6 ms -> 80.1 ms +- 0.7 ms: 1.01x faster
- pickle_list: 4.62 us +- 0.05 us -> 4.59 us +- 0.04 us: 1.01x faster
- xml_etree_iterparse: 106 ms +- 1 ms -> 105 ms +- 1 ms: 1.01x faster
- regex_compile: 143 ms +- 1 ms -> 142 ms +- 1 ms: 1.00x faster

gvanrossum · 2021-11-09T22:08:19Z

What exactly do the two columns represent?

It's the two results files that I was comparing.

I get that. :-)

What I didn't get is what is different between the two runs you are comparing. Were they two different commits on cpython's main branch? Were they using different versions of pyperformance? You write "my branch". Your branch of what?

Here's my current hypothesis. You were using two different versions of PyPerformance, i.e. the head of its main branch and the head of your branch (python/pyperformance#109), and you ran these with exactly the same cpython commit, and both with optimizations enabled. Moreover, the main PyPerformance branch is the left column and the PR 109 branch is on the right.

What did I miss?

ericsnowcurrently · 2021-11-09T22:14:33Z

You've described it exactly. :) I was checking to see if my changes impact the pyperformance results.

(Sorry about the confusion.)

ericsnowcurrently · 2021-11-10T00:55:35Z

It looks like a problem in Cython. See cython/cython#4461

The problem was fixed a couple weeks ago (for 0.29.x) but it's unclear when the next release (0.29.25) will happen: https://github.com/cython/cython/milestones.

Separately, it looks like uWSGI doesn't build currently on 3.11. I've opened unbit/uwsgi#2365.

gvanrossum · 2021-11-10T03:59:10Z

Do we need to wait for a Cython release or could we (temporarily) point our lockfile to a specific commit?

pablogsal · 2021-11-10T12:13:52Z

Do we need to wait for a Cython release or could we (temporarily) point our lockfile to a specific commit?

~~Unfortunately seems that the only realistic way that I can think of is a new Cython release as if we want to test new commits on top of main, that's the only way to get a working Cython.~~

Alternatively, we can try to install cython based on its master branch (IIRC pip allows a git-based specifier), but we need to make sure we can build the wheel ourselves (should happen automatically, but not sure what build dependencies and requirements cython has).

ericsnowcurrently · 2021-11-10T15:47:02Z

Alternatively, we can try to install cython based on its master branch (IIRC pip allows a git-based specifier)

I'll give it a shot.

ericsnowcurrently · 2021-11-10T19:03:06Z

The specifier is "Cython@https://github.com/cython/cython/archive/0.29.x.tar.gz".

I tried it and it looks like Cython is incompatible with main at the moment (PyCode_New(), PyFrameObject.f_localsplus, CO_NOFREE). 🙁

gvanrossum · 2021-11-10T19:29:40Z

Are there at least Cython issues that make the Cython team aware of those incompatibilities? Ideally we could probably submit PRs that fix them -- I think this is within the scope of our project.

I'm surprised that PyCode_New() is still an issue, I thought that Stefan had fixed this (cython/cython#4414).

ericsnowcurrently · 2021-11-10T20:15:36Z

Yeah, I'll check the Cython issue tracker and post a bug report if needed. I also don't mind putting up a fix for them if it isn't too hairy. 🙂

ericsnowcurrently · 2021-11-17T19:35:11Z

FTR, here are the issues for extension problems with 3.11 that I found while running the Pyston benchmarks against CPython's main:

gvanrossum · 2021-11-17T21:05:27Z

I think solving these is going to be a priority -- In #102 we are also running into this.

da-woods · 2021-12-06T18:27:27Z

Just FYI - I believe Stefan has released Cython 0.29.25 which should unblock the vast majority of Cython-based dependencies (at least until Python 3.11.a3...)

gvanrossum · 2021-12-06T18:31:43Z

Just FYI - I believe Stefan has released Cython 0.29.25 which should unblock the vast majority of Cython-based dependencies (at least until Python 3.11.a3...)

Well 3.11a3 comes out today, so let's hope it still works with that. We didn't change the code object creation API, but we did change other things since alpha 2.

ericsnowcurrently · 2022-02-08T16:52:42Z

Any remaining work is covered in follow-up tasks: #175, #253, #164.

ericsnowcurrently self-assigned this Jun 16, 2021

ericsnowcurrently added the blocked label Jul 28, 2021

ericsnowcurrently removed the blocked label Sep 1, 2021

This comment has been minimized.

Sign in to view

gvanrossum mentioned this issue Dec 8, 2021

Run the Pyston benchmarks and compare how we're doing with those versus CPython 3.10. #164

Closed

ericsnowcurrently added the blocked label Dec 13, 2021

ericsnowcurrently removed the blocked label Dec 13, 2021

gramster added this to Fancy CPython Board Jan 10, 2022

gramster moved this to In Review in Fancy CPython Board Jan 10, 2022

ericsnowcurrently mentioned this issue Jan 12, 2022

Get up-to-date results for the "Quantifying" research paper. #220

Open

7 tasks

ericsnowcurrently moved this from In Review to Done in Fancy CPython Board Jan 24, 2022

ericsnowcurrently moved this from Done to In Review in Fancy CPython Board Jan 24, 2022

ericsnowcurrently moved this from In Review to Done in Fancy CPython Board Feb 1, 2022

ericsnowcurrently moved this from Done to In Review in Fancy CPython Board Feb 1, 2022

ericsnowcurrently mentioned this issue Feb 1, 2022

Add pyston benchmarks to speed.python.org. #253

Open

ericsnowcurrently moved this from In Review to Done in Fancy CPython Board Feb 1, 2022

ericsnowcurrently closed this as completed Feb 8, 2022

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Pull pyston benchmarks into pyperformance. #60

Pull pyston benchmarks into pyperformance. #60

ericsnowcurrently commented Jun 16, 2021 •

edited

Loading

Fidget-Spinner commented Jun 18, 2021 •

edited

Loading

ericsnowcurrently commented Jul 23, 2021 •

edited

Loading

ericsnowcurrently commented Sep 1, 2021

This comment has been minimized.

ericsnowcurrently commented Nov 9, 2021

gvanrossum commented Nov 9, 2021

brandtbucher commented Nov 9, 2021

brandtbucher commented Nov 9, 2021 •

edited

Loading

ericsnowcurrently commented Nov 9, 2021

gvanrossum commented Nov 9, 2021 via email

ericsnowcurrently commented Nov 9, 2021

gvanrossum commented Nov 9, 2021

ericsnowcurrently commented Nov 9, 2021

ericsnowcurrently commented Nov 10, 2021 •

edited

Loading

gvanrossum commented Nov 10, 2021

pablogsal commented Nov 10, 2021 •

edited

Loading

ericsnowcurrently commented Nov 10, 2021

ericsnowcurrently commented Nov 10, 2021

gvanrossum commented Nov 10, 2021

ericsnowcurrently commented Nov 10, 2021

ericsnowcurrently commented Nov 17, 2021

gvanrossum commented Nov 17, 2021

da-woods commented Dec 6, 2021

gvanrossum commented Dec 6, 2021

ericsnowcurrently commented Feb 8, 2022

Pull pyston benchmarks into pyperformance. #60

Pull pyston benchmarks into pyperformance. #60

Comments

ericsnowcurrently commented Jun 16, 2021 • edited Loading

Fidget-Spinner commented Jun 18, 2021 • edited Loading

ericsnowcurrently commented Jul 23, 2021 • edited Loading

ericsnowcurrently commented Sep 1, 2021

This comment has been minimized.

ericsnowcurrently commented Nov 9, 2021

gvanrossum commented Nov 9, 2021

brandtbucher commented Nov 9, 2021

brandtbucher commented Nov 9, 2021 • edited Loading

ericsnowcurrently commented Nov 9, 2021

gvanrossum commented Nov 9, 2021 via email

ericsnowcurrently commented Nov 9, 2021

gvanrossum commented Nov 9, 2021

ericsnowcurrently commented Nov 9, 2021

ericsnowcurrently commented Nov 10, 2021 • edited Loading

gvanrossum commented Nov 10, 2021

pablogsal commented Nov 10, 2021 • edited Loading

ericsnowcurrently commented Nov 10, 2021

ericsnowcurrently commented Nov 10, 2021

gvanrossum commented Nov 10, 2021

ericsnowcurrently commented Nov 10, 2021

ericsnowcurrently commented Nov 17, 2021

gvanrossum commented Nov 17, 2021

da-woods commented Dec 6, 2021

gvanrossum commented Dec 6, 2021

ericsnowcurrently commented Feb 8, 2022

ericsnowcurrently commented Jun 16, 2021 •

edited

Loading

Fidget-Spinner commented Jun 18, 2021 •

edited

Loading

ericsnowcurrently commented Jul 23, 2021 •

edited

Loading

brandtbucher commented Nov 9, 2021 •

edited

Loading

ericsnowcurrently commented Nov 10, 2021 •

edited

Loading

pablogsal commented Nov 10, 2021 •

edited

Loading